Whitepaper

A GAME-CHANGING
APPROACH TO
STABLE CYCLING OF
LNMO-GRAPHITE CELLS

Authors:
Piotr Jankowski, Scientist
Christian Fink Elkjær, Cell Development Manager
Jonathan Højberg, R&D Director
Topsoe A/S, Copenhagen, Denmark
CONTENTS

INTRODUCTION 3

STATE OF CHARGE AND THE IMPACT OF BALANCING 4

AN LNMO-GRAPHITE CELL WITHOUT FADE 6

APPLICATION IN COMMERCIAL BATTERY CELLS 7

CONCLUSION 10

METHODS 11

Authors:

Piotr Jankowski, Christian Fink Jonathan Højberg,

Scientist Elkjær, Cell R&D Director
Development
Manager

Topsoe A/S, Copenhagen, Denmark

A game-changing approach to stable cycling of LNMO-Graphite Cells 2

INTRODUCTION

The need for cost-effective and high-performance into the SEI layer which increases impedance. This is
energy storage solutions has been a driving force in one of several phenomena generalized as “chemical
advancing battery technology. In this context, the crosstalk” where electrolyte-soluble species
combination of a high-voltage cathode material, produced on one electrode migrate to the other
like Lithium Nickel Manganese Oxide (LNMO) electrode and change cell resistance and/or lithium
with graphite as the anode, has emerged as an inventory, i.e., amount of active lithium.
exceptionally promising system for automotive
application. This paper will introduce “non-chemical crosstalk”
that relates to the interplay between state of charge
While LNMO-graphite cells present great promise at the electrodes which is largely governed by
for the future, insufficient capacity retention still N/P ratio and voltage window. Using 3-electrode
hinders their full commercialization. As reported in measurements, we show how optimization of these
literature1, the capacity retention is mainly limited by parameters drastically reduces the stress on each
the interplay between the graphite anode and the electrode and minimizes the chemical crosstalk,
LNMO cathode, often referred to as “crosstalk”. The resulting in significantly improved capacity retention.
most referred crosstalk is dissolution of Mn from the
cathode that migrates to the anode and integrates

A game-changing approach to stable cycling of LNMO-Graphite Cells TO CONTENTS 3

STATE OF CHARGE AND
THE IMPACT OF BALANCING

To understand the origin of non-chemical crosstalk, • The maintained constant current control in the
it is important to understand how the LNMO- LNMO cell until a higher state of charge leads to
graphite system is different compared to other drop of the anode voltage. At C/2 charge, this
common cathode chemistries. Figure 1 shows a causes the potential to reach below 0 V vs. Li/Li+.
comparison between half- and full-cell voltage This is not observed for the NMC cell.
profiles of LNMO-graphite and NMC532-graphite at
both slow (C/10, D/10) and fast (C/2, 1D) cycles. The In essence, LNMO cathode with flat potential profile
three main observations when comparing the two and easily accessible lithium cations allows to
systems are: maintain the CC charging step until even 98% SOC
at current rate C/2. Such conditions can easily lead
• Anode potential profiles are similar for both to early Li-plating on the anode electrode.
systems at low current conditions.
• The NMC cell reaches CV phase of charging
earlier than the LNMO cell. It is very pronounced
at fast cycling, but observable also for slow cycles.

FIGURE 1: Potential profile of LNMO-graphite and NMC-graphite cells

Potential / V LNMO: full cell NMC: full cell Potential / V LNMO: full cell NMC: full cell
LNMO: cathode NMC: cathode LNMO: cathode NMC: cathode
LNMO: graphite NMC: graphite LNMO: graphite NMC: graphite
5 5

4 4

0.35
3 3 0.30
0.25
0.20
2 2
0.15
0.10

1 1 0.05
0.00

0 0

0 2 4 6 8 10 12 14 16 0.0 0.5 1.0 1.5 2.0 2.5 3.0

Cycle time / h Cycle time / h

Left: Slow cycle (2nd cycle) at C/10 and D/10 and Right: Faster cycle (3rd cycle) at C/2 and 1D. Results are obtained using 3 electrode PAT-CELLs
(EL-CELL) with 2 cycles C/10 and D/10 followed by cycling at C/2 and 1D. Electrolyte: 1M LiPF6 EC:DEC (1:1 vol.) + 1% LiBOB. Balancing: N/P = 1.20.
Temperature: 25°C.

A game-changing approach to stable cycling of LNMO-Graphite Cells TO CONTENTS 4

Figure 2 shows typical evolution of anode and The increased lower potential of the cathode
cathode potentials upon cycling. Figure 2 left shows with cycle numbers reflects a decrease in lithium
anode and cathode potential profiles of cycle 5, inventory that prevents full lithiation of LNMO
25 and 45 and Figure 2 right shows maximum and and thereby less reduction of Mn4+ to Mn3+ at
minimum values of potential for both anode and around 4.3 V vs. Li/Li+. At a certain point, lithium
cathode during cycling. The three main observations inventory decreases to a level where no Mn4+ is
are: reduced. At this point, battery discharge ends with
reduction of Ni3+ to Ni2+ at around 4.68 V vs. Li/Li+.
• the potential of the anode drops below 0 V vs. As the lower cut-off voltage is 3.5 V, the increased
Li/Li+ from cycle 12 potential on the cathode leads to increasing anode
• the lower potential limit of the cathode at the potential reaching above 1 V vs. Li/Li+. As described
end of discharge increases with cycle numbers in literature2 one can speculate about how such
• the upper potential of the anode increases to high anode potentials can change the SEI-layer by
above 1 V vs. Li/Li+ from cycle 25 dissolution of some of its components, ultimately
lowering the protection ability of the SEI-layer.
With an anode potential below 0 V vs. Li/Li+, lithium
plating is thermodynamically possible and although In short, cycling of LNMO-graphite cells using the
the voltage curves are not sufficient to confirm same protocols as for NMC-graphite cells exposes
plating, it would eventually lead to fast decrease the anode to both very high and very low potentials.
of lithium inventory, i.e., loss of active lithium in the That sets demanding requirements to anode
system. material and may be the reason for limited capacity
retention of LNMO-graphite cells.

FIGURE 2: Potential profile of LNMO-graphite cell

Potential / V Cathode Potential / V Eanode: minimum Eanode: maximum

Anode Ecatode: minimum Ecatode: maximum
5 6

Cycle 5 Cycle 25 Cycle 45 5

4

4
3 Ecell=3.5V Ecell=3.5V

Ecell=3.5V 3
2
2

1
1

0 0
47 48 49 111 112 113 177 178 179
Cycle time /h Cycle time /h Cycle time /h -1
0 20 40 60 80 100
Cycle

Left potential profile for anode and cathode of LNMO-graphite cell in cycles 5, 25, and 45, and right maximum and minimum potentials experienced
by anode and cathode for each cycle. Results are obtained using 3 electrode PAT-CELLs (EL-CELL) with 2 cycles C/10 and D/10 followed by cycling at
C/2 and 1D. Electrolyte: 1M LiPF6 EC:DEC (1:1 vol.) + 1% LiBOB. Balancing: N/P = 1.20. Temperature: 25°C.

A game-changing approach to stable cycling of LNMO-Graphite Cells TO CONTENTS 5

AN LNMO-GRAPHITE CELL WITHOUT FADE

To test the impact of limiting stress on the anode, a

safe potential range for graphite anode was selected FIGURE 3: Cycling performance of LNMO-
as 0.1-0.8 V vs. Li/Li+, and LNMO-graphite cells were graphite cells with improved anode control
cycled with control of anode (anode potential vs.
lithium metal reference electrode). It is important to
understand the implications of such an approach: Discharge capacity / mAh/g N/P=1.64
N/P=1.10
130
• current flow between anode and cathode is
determined by the kinetics of the anode
120
• charging stops when the potential of the anode
reaches 0.1 V vs. Li/Li+
110
• discharging stops when the potential of the
anode reaches 0.8 V vs. Li/Li+
90
• the voltage and state of charge of the LNMO
cathode undergoes continuous adjustments as
80
cycling is determined by the anode

70
Limiting anode potential to 0.1-0.8 V vs. Li/Li+ limits
capacity of the anode with around 30%. To maintain
60
high capacity of the cell (in mAh/gLNMO), it is therefore
0 20 40 60 80 100 120
necessary to increase capacity of the anode. For the
graphite and LNMO used in this study, the optimal Cycle number

N/P ratio was 1.64. For comparison, a cell with N/

P=1.10 was also tested. Figure 3 shows cycling
Cycling performance of LNMO-graphite cell with anode control
curves of cells with the two N/P ratios. The results protocol (Eanode = 0.1-0.8 V vs. Li/Li+). Results are obtained using 3
demonstrate that it is possible to build stable electrode PAT-CELLs (EL-CELL). Electrolyte: 1M LiPF6 EC:DEC (1:1
vol.) + 1% LiBOB. Balancing: N/P = 1.10 and N/P = 1.64, respectively.
LNMO-graphite cells and maintain stability for more Temperature: 25°C.
than 100 cycles with a capacity of 120 mAh/gLNMO,
when the conditions of the cycling are governed
by the state of the anode. The stable cycling under
these conditions further indicates that the origin
of capacity loss in LNMO-graphite cells is linked
closely to non-chemical crosstalk, where the
combination of a standard CC-CV protocol and an
LNMO cathode forces extreme cycling conditions on
the anode, leading to the chemical crosstalk known
from literature.

A game-changing approach to stable cycling of LNMO-Graphite Cells TO CONTENTS 6

APPLICATION IN COMMERCIAL BATTERY CELLS

Monitoring anode potential during cycling is not Herein we would like to demonstrate the first step
feasible for commercial applications, and it is in that direction, obtained by reducing the upper
therefore necessary to translate the learnings from voltage cut-off potential and using an N/P ratio of
the 3-electrode system to normal cells without a 1.20. Figure 4 shows how the decrease in cut-off
reference electrode. potential will eventually compromise the available
capacity. The three selected upper cut-off potentials
Options to improve anode control could be are 4.80 V, 4.70 V, and 4.64 V. 4.80 V is selected
advanced cycling protocols, such as e.g., multistage as a reference often used in literature. 4.70 V is
constant current, pulse, or current modulated selected to decrease current towards end of charge
charging, to replicate potential and current profiles by entering CV step earlier and without significantly
obtained in anode-control mode in standard LNMO- compromise in capacity. 4.64 V is selected as the
graphite cells. best way to mimic results in the 3-electrode cell,
where the charge is terminated exactly as the anode
reaches 0.1 V vs. Li/Li+. As marked in Figure 4, a cut-
off voltage of 4.64 V in a cell with N/P ratio 1.2 will
FIGURE 4 : Charge potential profiles of LNMO be reached when the anode potential (difference
and full cell in LNMO-graphite cell between cathode potential and full cell potential)
changes from slightly above 0.12 V vs. Li/Li+ to
around 0.08 V vs. Li/Li+ as the cathode potential is
4.74 V vs. Li/Li+ in this region.
Potential / V Full cell / V
Cathode / V
An important effect of changing the upper voltage
4.85 cut-off is that the current profile towards end of
Ecell = 4.80 V charge is modified significantly. Figure 5 shows
4.80 capacity vs. time for the three conditions. The slope
4.75
corresponds to the current, and the CV step is
initiated when the slope decreases from the initial
Ecell = 4.70 V
4.70 straight line. As further substantiated in Table 1,
0.1 V
the CC charging step is reduced from 93% down
4.65 Ecell = 4.64 V
to 63%. This means that lowering the upper cutoff
4.60
voltage decreases the available capacity and
increases time at CV from 16% to 53%. The total
4.55 charging time is almost unchanged as the capacity
is 20% less for the upper cut-off voltage of 4.64 V
0 25 50 75 100 125
compared to 4.80 V.
Capacity / mAh/g

Results are obtained using 3 electrode PAT-CELLs (EL-CELL) at

C/10 charge. Electrolyte: 1M LiPF6 EC:DEC (1:1 vol.) + 1% LiBOB.
Balancing: N/P = 1.20. Temperature: 25°C.

A game-changing approach to stable cycling of LNMO-Graphite Cells TO CONTENTS 7

 FIGURE 5: Impact of upper cut-off FIGURE 6: Cycling performance based
voltage on current and capacity on full range capacity

Capacity / mAh/g 3.50-4.80 V 3.50-4.64 V Full range capacity / mAh/g 3.50-4.80 V 3.50-4.64 V
Full range capacity / mAh/g 3.50-4.80 V
3.50-4.70 V 3.50-4.64 V
3.50-4.70 V
3.50-4.70 V
120 125
125

100 120
120

80 115
115

60 110
110

40 105
105

20 100
100

95
0 95
0 20 40 60 80 100 120 140 160
0.00 0.25 0.50 0.75 1.00 1.25 1.50 0 20 40 60 80 100 120 140 160
Cycle number
Time / h Cycle number

Charge at C/2 with varying upper cut-off voltage of 4.80 V, 4.70 Results are obtained in 2 electrode coin cells cycled at C/2 and
V and 4.64 V, respectively. Results are obtained in 2 electrode 1D with varying upper cut-off voltage of 4.80 V, 4.70 V and 4.64
coin cells. Electrolyte: 1M LiPF6 EC:DEC (1:1 vol.) + 1% LiBOB. V, respectively. Full range capacity is measured every 50th cycle
Balancing: N/P = 1.20. Temperature: 25°C. between 3.50 V and 4.80 V at C/10 and D/10. Electrolyte: 1M LiPF6
EC:DEC (1:1 vol.) + 1% LiBOB. Balancing: N/P = 1.20. Temperature:
25°C.

TABEL 1: Key values for charge at C/2 with To provide periodical ‘check’ of the available
different upper cut-off voltages capacity, the cells were cycled slowly (C/10, D/10)
in the full voltage range of 3.5 V–4.8 V every 50
Potential Qtotal QCC/ ttotal tCV/ cycles. Figure 6 shows capacities at slow cycles of
range [V] [mAh/g] Qtotal [h] ttotal the three conditions and comparing these, it is clear
that milder cycling significantly improves capacity
3.50-4.80 110.9 93% 1.66 16% retention. This result is further substantiated in Table
2 that shows capacities and capacity retention
3.50-4.70 107.0 85% 1.64 25% corresponding to the three conditions. On one hand,
the available capacity is lower when the upper cut-
3.50-4.64 87.0 63% 1.61 53% off is decreased, while at the same time the capacity
retention is much better, decreasing the capacity
Results are obtained in 2 electrode coin cells at C/2 (5th cycle) fade at slow cycles and full voltage range from 5.6
with varying upper cut-off voltage of 4.80 V, 4.70 V and 4.64 mAh/g to 1.5 mAh/g in the first 50 cycles. Loss of
V, respectively. Electrolyte: 1M LiPF6 EC:DEC (1:1 vol.) + 1% LiBOB.
Balancing: N/P = 1.20. Temperature: 25°C. capacity speeds up in subsequent cycle ranges, but
is always the slowest when lower cut-off voltage is
applied.

A game-changing approach to stable cycling of LNMO-Graphite Cells TO CONTENTS 8

 TABEL 2: Key values for cycling with different upper cut-off voltages

Potential range Qfull (cycle 1) Qlimited (cycle 5) Capacity loss cycles Capacity loss cycles Capacity loss cycles
[V] [mAh/g] [mAh/g] 2-52 52-102 102-152
[mAh/g] [mAh/g] [mAh/g]

3.50-4.80 120 110 5.6 5.7 3.6

3.50-4.70 121 106 5.0 3.7 3.5

3.50-4.64 120 87 1.5 3.5 3.3

Results are obtained in 2 electrode coin cells with varying upper cut-off voltage of 4.80 V, 4.70 V and 4.64 V, respectively.
Electrolyte: 1M LiPF6 EC:DEC (1:1 vol.) + 1% LiBOB. Balancing: N/P = 1.20. Temperature: 25°C.

To better understand how the voltage window FIGURE 7: Cell resistance during cycling
impact the cell, intermittent current interruption (ICI)
method was used to monitor cell resistance.
Increase in cell resistance / ohm 3.50-4.80 V 3.50-4.64 V
3.50-4.70 V
Figure 7 shows ICI measurements conducted
40
during test of the three conditions and reveals that
resistance build-up is much more significant at high
upper voltage cut-off, while the cell cycled to 4.64 30
V did not show any noticeable increase in resistance
for 100 cycles. After cycle 102 performed in full range 20
3.50–4.80 V, the build-up in resistance speeds up.

10
Thus, we can only assume that, even a single
slow cycling in full range causes some irreversible
changes in the cell, and the overall performance of 0
the cell would be better if these cycles were skipped.

0 20 40 60 80 100 120 140 160

Cycle number

Resistance is determined using ICI methodology. Results are

obtained in 2 electrode coin cells with varying upper cut-off
voltage of 4.80 V, 4.70 V and 4.64 V, respectively. Electrolyte:
1M LiPF6 EC:DEC (1:1 vol.) + 1% LiBOB. Balancing: N/P = 1.20.
Temperature: 25°C.

A game-changing approach to stable cycling of LNMO-Graphite Cells TO CONTENTS 9

CONCLUSION

In conclusions, our research has demonstrated

that it is possible to create a stable LNMO-
graphite cell configuration with essentially no
degradation. The key to achieving this is controlling
the potential range of the anode during cell
cycling. Comparison between NMC and LNMO
cathodes shows that potentials below 0 V vs. Li/
Li+ on the anode is much more likely with LNMO
due to specific intrinsic properties of LNMO and
that optimization of the cycling protocol is key
to avoid this. While the approach presented here
aims to provide proof-of-concept, it is important
to acknowledge that it is not perfect and requires
further development to maintain the observed low
initial capacity degradation. Our initial trials have
shown that reducing capacity fading is possible,
which is a promising step towards developing high-
performance energy storage solutions based on
LNMO-graphite cells.

A game-changing approach to stable cycling of LNMO-Graphite Cells TO CONTENTS 10

METHODS

LNMO electrodes with loading of around 1 mAh/cm2 2-electrode measurements were performed in coin-
were prepared from TOPSOE LNMO material, using cell geometry, using Celgard H2010 separator and
procedures described in previous publications3. 20 µL electrolyte: 1M LiPF6 in EC:DEC (vol. 1:1) + 1%
LiBOB. Cycling was performed at stable 25 °C:
3-electrode measurements were performed using
PAT-CELL provided by EL-CELL®, using Freudenberg • First two cycles C/10, D/10
Viledon FS 2226E + Lydall Solupor 5P09B separator • Subsequent cycles were performed using C/2
and 100 µL electrolyte: 1M LiPF6 in EC:DEC (vol. 1:1) + (+ CV until current below C/10), 1D
1% LiBOB. Cycling was performed at stable 25 °C: • Every 50 cycle, the slow cycle was performed:
C/10, D/10
• First two cycles C/10 (+ CV until current below Potential range for slow cycles was set to
C/20), D/10 (+ CV until current below D/20); deep 3.50–4.80 V for all cells, and fast cycles were
stabilization of current was done to perform performed in range 3.50–4.80 V, 3.50–4.70 V
Electrochemical Impedance Spectroscopy at or 3.50–4.64 V
both charged and discharged state • The cell resistances were monitored during
• Subsequent cycles were performed using C/2 discharge of each cycle using ICI procedure4
(+ CV until current below C/20), 1D (+ CV until
current below D/20)
• Potential range was set to 3.50–4.80 V for LNMO-
graphite (cell control), 3.00–4.40 V for NMC532-
graphite (cell control), 0.80–0.10 V for LNMO-
graphite (anode control)

REFERENCES

1
https://doi.org/10.1016/j.electacta.2012.12.069
2
https://doi.org/10.1002/aenm.201801427
3
https://www.topsoe.com/hubfs/Topsoe_status_LNMO.pdf
4
https://doi.org/10.1016/j.electacta.2022.140888

A game-changing approach to stable cycling of LNMO-Graphite Cells TO CONTENTS 11

Founded in 1940, Topsoe is a global leader in
developing solutions for a decarbonized world,
supplying technology, catalysts, and services for
worldwide energy transition.

Our mission is to combat climate change by

helping our partners and customers achieve their
decarbonization and emission-reduction targets,
including those in challenging sectors such as aviation,
shipping, and production of crucial raw materials.

From low-carbon or zero-carbon chemicals to

renewable fuels and plastic upcycling, we are uniquely
positioned to aid humanity in realizing a sustainable
future.

Topsoe is headquartered in Denmark, with 2,100

employees serving customers all around the globe.
To learn more, visit topsoe.com

Get in touch today

topsoe.com/contact

Topsoe A/S,
Haldor Topsøes Allé 1
2800 Kongens Lyngby
Denmark

Tel. +45 45 27 20 00
CVR no. 41 85 38 16

0389.2023/Rev.0