Decoupling Interfacial Reactions at Anode and Cathode by Combining Online Electrochemical Mass Spectroscopy with Anode-Free Li-Metal Battery

Article abstract of DOI:10.1002/adfm.202006951

It is essential to decouple the interfacial reactions?taking?place at the anode and cathode in?rechargeable batteries. However, due to the reactive nature of Li, it is challenging to use Li-metal batteries (LMBs) protocol to decouple the interfacial reactions. The?by-products from the anode or cathode become mixed in Li/NMC111 cells, which make decoupling interfacial reactions difficult. Here, reactions at electrodes are successfully decoupled and demystified using a protocol combining anode-free LMB (AFLMB) with online electrochemical mass spectroscopy. LiPF₆?in ethylene carbonate (EC)/diethyl carbonate (DEC) and EC/ethyl methyl carbonate (1:1 v/v%) electrolytes are used to compare interfacial reactions in Li/NMC111 and Cu/NMC111 cells. In Cu/NMC111, the evolution of CO₂, CO, and C₂H₄?gases at the initial stage of first charging is due to interfacial reactions at Cu surface due to solid–electrolyte-interphase formation. However, the evolution of CO₂?and CO gases at high voltage in the entire cycles is associated with chemical and/or electrochemical electrolyte oxidation at the cathode. This work paves a new concept to decouple interfacial reactions at electrodes for developing electrochemically stable electrolytes to improve the performance with the long-cycling life of AFLMBs and LMBs.

Full text of DOI:10.1002/adfm.202006951

Full PaPer
www.afm-journal.de
Decoupling Interfacial Reactions at Anode and Cathode
by Combining Online Electrochemical Mass Spectroscopy
with Anode-Free Li-Metal Battery
Kassie Nigus Shitaw, Sheng-Chiang Yang, Shi-Kai Jiang, Chen-Jui Huang, Niguse Aweke Sahalie,
Yosef Nikodimos, Haile Hisho Weldeyohannes, Chia-Hsin Wang, She-Huang Wu,*
Wei-Nien Su,* and Bing Joe Hwang*
hydrogen electrode (SHE)) of Li-metal.[1]
It is essential to decouple the interfacial reactions taking place at the anode
and cathode in rechargeable batteries. However, due to the reactive nature
of Li, it is challenging to use Li-metal batteries (LMBs) protocol to decouple
the interfacial reactions. The by-products from the anode or cathode become
mixed in Li/NMC111 cells, which make decoupling interfacial reactions dif-
ficult. Here, reactions at electrodes are successfully decoupled and demystified
using a protocol combining anode-free LMB (AFLMB) with online electrochem-
ical mass spectroscopy. LiPF6 in ethylene carbonate (EC)/diethyl carbonate
(DEC) and EC/ethyl methyl carbonate (1:1 v/v%) electrolytes are used to com-
pare interfacial reactions in Li/NMC111 and Cu/NMC111 cells. In Cu/NMC111,
the evolution of CO2, CO, and C2H4 gases at the initial stage of first charging
is due to interfacial reactions at Cu surface due to solid–electrolyte-interphase
formation. However, the evolution of CO2 and CO gases at high voltage in the
entire cycles is associated with chemical and/or electrochemical electrolyte
oxidation at the cathode. This work paves a new concept to decouple interfa-
cial reactions at electrodes for developing electrochemically stable electrolytes
to improve the performance with the long-cycling life of AFLMBs and LMBs.
However, Li-metal anode is aggressively
reactive with carbonate-based electrolytes,
resulting in electrolyte consumption, and
depletion of active lithium (Li).[2] As a
result, LMBs suffered from low coulombic
efficiency, short cycling life, and short cir-
cuit especially at elevated temperatures,
higher voltage, and fast cycling rates. For
carbonate electrolyte based batteries, the
decomposition of electrolyte causes the
evolution of gases that lead to the expan-
sion of the cell.[3,4] The evolution of gases
is one of the indications of the instability
of electrolytes due to the electrode–elec-
trolyte interfacial reactions. Different
approaches designed to investigate the
gas evolution along with interfacial reac-
tions include Archimedes’ in situ gas
analyzer[5] and ex situ GC-MS[6] via LMBs
and Li-ion batteries (LIBs). However, these
approaches are not capable of showing the
reaction stages and conditions at which
gaseous by-products start to evolve in the cell. Recently, many
research groups have been using online electrochemical mass
spectroscopy (OEMS) approach for the study of gas evolution
along with interfacial reactions in the rechargeable LMBs and
LIBs.[6–9] It has been reported that the interfacial reactions upon
charging of Li/NMC and graphite/NMC cells with carbonate
electrolytes generate CO2, CO, H2 and C2H4 gases.[3,7,9–13]
1. Introduction
Rechargeable lithium metal batteries (LMBs) is an attractive
next-generation energy storage device for consumer electronics,
electric vehicles (EVs), and smart grid energy storage due to the
high theoretical specific capacity (3860 mAh g−1) and lowest
negative electrochemical potential (−3.04 V versus standard
K. N. Shitaw, Dr. S.-C. Yang, S.-K. Jiang, C.-J. Huang, N. A. Sahalie,
Y. Nikodimos, H. H. Weldeyohannes, Prof. B. J. Hwang
NanoElectrochemistry Laboratory
C.-H. Wang, Prof. B. J. Hwang
National Synchrotron Radiation Research Center
Hsinchu 30076, Taiwan
Prof. S.-H. Wu, Prof. W.-N. Su
NanoElectrochemistry Laboratory
Graduate Institute of Applied Science and Technology
National Taiwan University of Science and Technology
Taipei 10607, Taiwan
Department of Chemical Engineering
National Taiwan University of Science and Technology
Taipei 10607, Taiwan
E-mail: bjh@mail.ntust.edu.tw
Dr. S.-C. Yang
Department of Chemical and Materials Engineering
Lunghwa University of Science and Technology
Taoyuan 33306, Taiwan
E-mail: wush@mail.ntust.edu.tw, wush88@gmail.com;
wsu@mail.ntust.edu.tw
Prof. S.-H. Wu, Prof. W.-N. Su, Prof. B. J. Hwang
Sustainable Energy Development Center
National Taiwan University of Science and Technology
Taipei 10607, Taiwan
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/adfm.202006951.
DOI: 10.1002/adfm.202006951
Adv. Funct. Mater. 2020, 2006951
2006951 (1 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
Nevertheless, the main source of hydrogen (H2) gas at the early
stage of solid–electrolyte-interphase (SEI) formation is the
reduction of residual trace H2O at the anode surface with the
formula of H2O + e− → OH− + ½ H2.[7,8,14]
decomposition pathways those lead to gas evolution in AFLMBs
and LMBs. The proposed protocol and reaction mechanisms
convey a new concept of interfacial reactions that take place at
both electrodes separately for developing chemically and elec-
trochemically stable electrolytes in the enhancement of perfor-
mance and cycling life of rechargeable batteries.
One of the approaches to improve the performance of
rechargeable batteries is developing chemically and electro-
chemically stable electrolyte against lithium metal (Li). Avoiding
interfacial reactions at the electrode–electrolyte interface grasp
the key for developing stable electrolytes for high-performance
batteries.[15] Therefore, it needs to have fundamental under-
standings of the interfacial reactions at the anode and cathode
separately. However, due to the highly reactive nature of Li
anode with electrolytes, the gaseous by-products from the elec-
trode–electrolyte interfacial reactions become mixing and cross-
talk effects occur, which makes difficult to decouple the anode
and cathode interfacial reactions using Li anode in LMBs. As
a result, it is difficult to demystify the initial interfacial reac-
tions in LMBs since it is problematic to differentiate the inter-
facial reactions at both electrodes. Under these circumstances,
anode-free cell configuration (Cu/NMC111) in which bare Cu-
foil used as anode current collector, can serve as a powerful tool
to decouple the interfacial reactions takes place at anode and
cathode.
Herein, the interfacial reactions at both electrodes are
decoupled and the initial interfacial reactions have been suc-
cessfully demystified using a protocol combining online gas
evolution analysis with anode-free lithium metal batteries
(Cu/NMC111) cycling at 25, 40, or 60 °C. The results were
compared for two electrolytes, LiPF6 in a mixture of ethylene
carbonate (EC)/diethyl carbonate (DEC) (1:1 v/v%) and eth-
ylene carbonate (EC)/ethyl methyl carbonate (EMC) (1:1 v/v%).
Based on the SEI film composition analysis with X-ray photo-
electron spectroscopy (XPS) and gas formation analysis from
online electrochemical mass spectroscopy (OEMS), we have
proposed the proper initial stage and high voltage electrolyte
2. Results and Discussion
2.1. Gas Formation Analysis of Cu/NMC111 Full-Cells at 25 °C
To demystify the initial interfacial reactions, the anode and
cathode surface reactions were decoupled using anode-
free Cu/NMC111 cells combined with OEMS protocol. The
Cu/NMC111 full-cells were assembled by following steps
described in Figure 1 and Figure S1, Supporting Information
with EL1 (1 m LiPF6 EC/DEC) or EL2 (1 m LiPF6 EC/EMC) elec-
trolytes and connected to OEMS.
The evolved gases were then measured online by using
OEMS at open-circuit voltage (OCV) followed by during cycling
at a cell temperature of 25, 40, or 60 °C. Since gases are evolved
at the first few formation cycles due to SEI and cathode–
electrolyte-interface (CEI) formation at anode and cathode,
respectively,[12,16,17] the evolution of gases were measured for
the first three cycles. Figure 2 depicts the potential versus time
profiles (upper panels) of the first three cycles of Cu/NMC111
full-cells with EL1 and EL2 electrolytes cycled at 25 °C, and
the total ion signals of evolved carbon dioxide (CO2), carbon
monoxide (CO), oxygen (O2), hydrogen (H2), ethylene (C2H4),
ethane (C2H6), and POF3 (lower panels). Different researchers
reported their outlooks about the sources of gases before O2
releasing in the battery cells. He et al.[8] and Bi et al.[18] reported
in their independent works that the origin of CO2 before O2
releasing in Li/NMC is from Li2CO3 residual neutralization on
Figure 1. Schematic illustration of the EL-CELL connecting with GC-MS designed in this work for online gas formation analysis in the AFLMBs and
LMBs. The Li-metal foil was used instead of the Cu substrate for LMBs configuration.
Adv. Funct. Mater. 2020, 2006951
2006951 (2 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
Figure 2. Variation of cell voltage (upper panel) and evolved gases (lower panel) versus time profile of Cu/NMC111 prepared with 250 µL of a) EL1,
b) EL2, respectively during cycling at 1 mA cm−2 and 25 °C within 2.5–4.8 V voltage range. The OEMS data were smoothed.
the surface of the cathode. In the Cu/NMC111 full-cells shown
in Figure 2a,b for EL1 and EL2 electrolytes, respectively, the
interfacial reactions take place at the Cu current collector and
cathode (NMC111) surfaces are decoupled by means of gas evo-
lution test occurred at the two electrodes. In Figure 2, CO2 gas
evolves steeply while CO and C2H4 gases increase stepwise at
the initial stage of the first cycle charging for both electrolytes.
The evolution of CO2, CO, and C2H4 gases at the initial stage of
first cycle charging is originated from the interfacial reactions
due to electrochemical electrolyte reduction during SEI forma-
tion on Cu substrate and is unaffected by the CEI reactions.
In Figure 2a,b, the evolution of gases at the initial stage of
first charging is associated with the electrochemical reduc-
tion of electrolyte solvents on the Cu substrate. The evolution
of C2H4 reductive gas simultaneous with CO2 and CO gases
starting from initial stage upon charging suggests the occur-
rence of electrochemical solvents reduction reactions at the Cu
surface in Cu/NMC111 cells. For EC solvent based electrolytes,
EC has main role on SEI formation due to its early reduction
at lower voltage.[19,20] EC undergoes ring-opening reduction
reaction followed by dimerization and elimination of ethylene
gas to produce Li2EDC as SEI component.[16] Pritzl et al.[21]
also figure out the mechanism of EC reduction which provides
the CO2 and CO gaseous products. This is because, due to its
higher electron affinity nature than linear carbonates (DEC
and EMC), EC solvent undergoes electrochemical reduction
reaction for SEI formation[22,23], suggesting that the evolution
of gases in batteries with in EC based electrolytes preferen-
tially associated with EC decomposition. Scheme 1 shows the
possible electrochemical interfacial reduction pathways of EC
solvent on the anode surface, mainly responsible for the evolu-
tion of CO2, CO and C2H4 reductive gases at the Cu current
collector.[24,25] However, the insignificance of H2 reductive gas
evolution in Cu/NMC111 cells indicates the main source of H2
is moisture in the batteries.
The contribution of EC solvent in EL1 and EL2 electrolytes
for the evolution of CO2, CO, and C2H4 gases in Cu/NMC111
full-cells has been more evidenced by conducting OEMS anal-
ysis for the analogous cell using 1 m LiPF6 in EC solvent as an
electrolyte (Figure S2, Supporting Information). To evaluate the
Scheme 1. Possible ring-opening reduction pathways consistent with OEMS and XPS results for EC solvent: I) Single EC solvent reduction pathways
and II) EC solvent dimerization (Li2EDC formation and its decomposition in SEI).
Adv. Funct. Mater. 2020, 2006951
2006951 (3 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
Scheme 3. Possible electrochemical oxidation pathways of EC solvent
consistent with the OEMS results at the cathode side.
Therefore, in Figure 2a,b, the sources of CO2 and CO gases in the
first cycle near to 4.39 and 4.38 V for EL1 and EL2 electrolytes,
respectively as well as in the second and third cycles are from
the interfacial reactions at the cathode side due to i) chemical
electrolyte oxidation via 1O2 (Equation 3) and ii) electrochemical
electrolyte oxidation as shown in Scheme 3. For Cu/NMC111
full-cells with EL1 or EL2 electrolytes, the molecular O2 gas is
generated in the first cycle charge due to the NMC111 cathode
degradation as shown in Equation (2), in good agreement with
reported in the literatures.[10,27,28]
Scheme 2. Proposed electrochemical reduction pathways of EMC and
DEC solvents for the evolution of gases detected herein. R1 and R2 can


be CH3 and/or C2H5.
role of EC ring-opening reduction for gas evolution, the cycling
protocol was unchanged. As depicted in Figure S2, Supporting
Information, the steepest CO2 evolution at the initial stage indi-
cates the ring-opening reduction of EC solvent (Scheme 1) due
to EC-derived SEI formation. The evolution intensity of CO2 in
Cu/NMC111 full-cell with EL1 and EL2 electrolytes shown in
Figure 2 is lower than obtained in the cell containing 1M LiPF6
in EC solvent only, suggesting that the electrochemical stability
of EL1 and EL2 electrolytes is higher than solvents alone due
to the synergetic effect of solvents. However, since DEC and
EMC solvents are very volatile during purged with helium car-
rier gas, the analogous OEMS gas formation analysis from cells
containing DEC and EMC solvents alone is not reported herein.
Reduction of linear carbonate solvents also have their own effect
on the evolution of CO2 and CO gases and other solid products.
Scheme 2 shows the possible electrochemical reduction path-
ways of EMC and DEC solvents.[24] In Scheme 2, the evolution
LiNi1/3Mn1/3Co1/3O2 → LiNi1/3Mn1/3Co1/3O2−x s +1/2x O
( )
g
(2)
2 ( )
In the cells, with NMC111 as cathode, singlet oxygen (1O2)
is released from the degradation of active materials at or above
4.3 V via Equation 2. The singlet oxygen (1O2) then undergoes
immediate chemical reaction with electrolyte solvents mainly
EC solvent (C3H4O3), leading to generation of CO2 and CO
gases via Equation 3.[19,29] As a result, the evolution of CO2 and
CO gases at the same time with O2 evolution suggesting that
the origin of CO2 and CO is from the chemical oxidation reac-
tion of electrolyte.
C3H4O3 +1 O2 g → 2CO g +CO g +2H O l
( )
(3)
( )
2 ( )
( )
2
The released trace H2O (moisture) in Equation (3) simulta-
neous with CO2 and CO gases causes hydrolysis of LiPF6 salt
in the electrolyte, which is evidenced by POF3 evolution via
Equation (4).

of CO gas occurs only when the carbonyl carbon oxygen bonds
(bonds II and III) is cleaved as shown path II/III in Scheme 2.

The C O bond in EMC and DEC structures (bonds I and
IV) is easily breakdown when the Li+ attacks the O atom of the
C O bond.[26] In the reduction pathways I and IV, the R1 and/or

LiPF s +H O l ꢀ POF g +LiF s +PF g
( )
6 ( ) ( )
(4)
3 ( )
5 ( )
2
C H.
CH.
)
3
R2 free radicals assigned for ethyl (
6) or methyl (
2
radical those capable to undergo proton transfer reaction and
The POF3 and/or PF5 further catalyze electrolyte decomposi-
tion, leading to evolving CO2 and CO gases. The C2H4 gas is
not detected in the second and third cycles for both EL1 and
EL2 containing cells, indicating that the main source of the CO2
and CO gases in the subsequent cycles is the chemical and/or
electrochemical electrolyte oxidation reactions at the NMC111
surface. Carbonate electrolytes also electrochemically oxidize
at the cathode surface above 4.3 V and causes to release gases
such as CO2 and CO at the cathode side.[13] As a result, the
evolved gases due to chemical and electrochemical oxidation
of electrolytes at the cathode side cannot be decoupled easily.
However, the evolution of O2 is insignificant in the second and
lead to generate C2H6 or CH4 gas as shown in Equation (1).[19]
R1/R2 +H+ +e− → C2H6 g /CH
g
4 ( )
(1)
( )
However, the R1OH3LiO2 and R2OH3LiO2 species fur-
ther react with H2O, HF or Li-metal to produces CO2 gas.
Nevertheless, the CH4 signal is insignificant during OEMS gas
formation analysis, indicating that the CH3 radical formation
reaction is minor. In the Cu/NMC111 cells, the electrolyte oxida-
tion interfacial reactions take place at a higher voltage is the main
sources of CO2 and CO gases at the cathode (NMC111) side.[10]
Adv. Funct. Mater. 2020, 2006951
2006951 (4 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
third cycles, and cannot enough to be detectable. Therefore, the
electrochemical oxidation of electrolyte solvents (mainly EC sol-
vent) is the major source for the evolution of CO2 and CO gases
at high voltages in the second and third cycles.[30] Scheme 3
shows the electrochemical EC solvent oxidation pathways
occurred at cathode surface. However, pathway I is assumed
to be the major process of electrochemical EC oxidation which
produces CO gas due to the presence of unstable oxygen radical
(O.) in the ring-opened EC.[17,27]
The rate of CO2 and CO gases evolution decreases from cycle
to cycle once the SEI is formed in the first formation cycle.
However, CO2 is detected until near to the end of the first dis-
charge of Cu/NMC111 with EL1 (Figure 2a). This is due to the
slow column flow rate of accumulated CO2 with a carrier gas
(0.45 mL min−1). Schwenke et al.[31] found that more amount
of CO gas was evolved from CV test of graphite/LFP cell con-
taining 1 m LiPF6 EMC as electrolyte. The Cu/NMC111 cell with
EMC containing EL2 electrolyte also shows relatively higher
CO intensity than other released gases, suggesting that EMC is
more electrochemically reductive solvent than DEC to produces
CO, especially at elevated temperature.[24] The main evidence
for the evolution of CO from EMC reduction is the release of a
large amount of C2H4 at the beginning of the first cycle at 25 °C
for the cell with EL2 than with EL1.
salt in the electrolyte produces strong Lewis acid (PF5) which
further reacts with moisture and produces POF3 (m/z = 85) and
HF (m/z = 20) as shown in Equations (5) and (6).[32]
In Figure S4, Supporting Information, the evolution trends of
POF3 and POF2 is similar even though the evolution of POF2 is
relatively higher than POF3 evolution. Therefore, the evolution of
POF3 is characterized by POF2 evolution which is the fragment
of POF3. The evolution of POF3 in the second and third cycles of
Cu/NMC111 cells at 40 or 60 °C indicates the direct reaction of
H2O with PF5.[11]
LiPF s →LiF s +PF g
6 ( ) ( )
5 ( )
∆
(5)
(6)
PF g +H O l →POF g +2HF g
( )
( )
5 ( )
3 ( )
2
The POF3 and/or PF5 catalyze the decomposition of organic
solvents (EC, DEC, and EMC) to produce CO2 gas[33] as shown
in Equations (7) and (8) and inorganic SEI components such
as LiF and Li2CO3 on the Cu current collector.[23] However, the
acidic species HF attacks the SEI, leading to further interfacial
reaction at the Cu surface and cause the evolution of CO2 and
CO gases via Schemes 1 and 2.[34]
POF /PF5
3
For the prevention of electrolyte and cathode materials from
decomposition at cathode surface, the operation cutoff voltage
should not exceed the normal operation voltage range. As
shown in Figure S3a,b, Supporting Information, no CO2 and
CO are evolved in the second cycle of Cu/NMC111 cells in
2.5–4.3 V voltage range for EL1 and EL2 electrolytes.
nC H O → C H O +CO2
g
(7)
(8)
(
)
( )
3
4
3
2
4
n
R1CO3R2 POF3/PF5→OR1 +R2 +CO2
g
( )
To explore the origin of O2 gas at the beginning of first
charging of Cu/NMC111 with EL1 or EL2 electrolytes (Figure 3)
at 40 or 60 °C, the evolution of gases were analyzed from
thermal decomposition of EL1 and EL2 in the absence of Cu,
NMC111 and separator (Figure S5, Supporting Information).
The evolution of O2 gas in Figure S5, Supporting Information
is associated with thermal decomposition of electrolyte sol-
vents, suggesting that the evolution of O2 at the initial stage
Figure 3a–d is originated from thermal electrolyte decomposi-
tion; not from cathode decomposition. Therefore, the very steep
increase of the CO2 and CO signals at the initial stages of the
first cycle at 40 and 60 °C also caused by chemical electrolyte
oxidation via Equation 3. Besides, the higher intensity of CO
than CO2 in Figure S5, Supporting Information, indicates CO
could be derived mainly from chemical electrolyte reactions,
and the kinetics of CO releasing reactions are highly boosted by
temperature. The O2, CO2, and C2H4 gases also released from
thermal decomposition of SEI components such as dilithium
ethylene dicarbonate (Li2EDC),[12,35] as shown in Scheme 1. The
evolution of CO2 and CO gases at the same time indicates that
the two gases have the same sources.[31,36] Therefore, the evolu-
tion of CO2 and CO gases from all Cu/NMC111 cells (Figures 2
and 3) and the consumption with EL1 and EL2 electrolytes coin-
cided, suggesting that CO2 and CO gases are from the same
sources in the cells. So it is unlikely that CO2 is fragmented to
CO by-product. It is reported that the EC-derived SEI at the
anode can be expanded and fractured upon charging at high
voltage (>4.3 V) and temperature.[27] In good agreement, the
C2H4 signal in the second and third cycles of Cu/NMC111 cells
2.2. Gas Formation Analysis of Cu/NMC111 Full-Cells
at 40 and 60 °C
To evaluate the effect of temperature on the reaction kinetics
and rate of gas evolution, the Cu/NMC111 full-cells were assem-
bled using EL1 or EL2 electrolytes separately. The cells then
were cycled at 40 or 60 °C and the evolved gases were measured
using the established protocol. Figure 3 shows the decoupling
of interfacial reactions at Cu current collector and NMC111
cathode of Cu/NMC111 cells cycled at 40 or 60 °C. The inten-
sity of CO2, CO, C2H4, and C2H6 gases especially at the initial
stage of first cycle charging at 40 or 60 °C is very intensive than
obtained at 25 °C (Figure 2). This shows that the kinetics of
SEI forming interfacial reaction at the Cu current collector is
boosted by elevated temperature. However, the evolution of
CO2 and CO gases at the same time with O2 at voltages of 4.36,
4.37, 4.35, and 4.36 V as shown in Figure 3a–d, respectively is
associated with chemical and/or electrochemical electrolyte
interfacial reactions at the cathode (NMC111) surface and trig-
gered by temperature. However, the intensity of gases in the
first cycle charging at 60 °C (Figure 3c,d) for EL1 and EL2 elec-
trolytes, respectively decreased quickly due to the conversion of
gases to other products. For example, CO2 can be electrochemi-
cally reduced to lithium oxalate (LiCO2)2 at the Cu surface.[31]
Similarly, the evolution of CO2 and CO gases in the second and
third cycles is originated from the electrochemical electrolyte
oxidation at the NMC111. The thermal decomposition of LiPF6
Adv. Funct. Mater. 2020, 2006951
2006951 (5 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
Figure 3. Variation of cell voltage (upper panel) and evolved gases (lower panel) versus time profile of Cu/NMC111 full-cells with 250 µL of a, c) EL1
and b,d) EL2 electrolytes, respectively during cycling at 40 or 60 °C within 2.5–4.8 V voltage range at 1 mA cm−2. The OEMS data were smoothed.
cycled at 40 or 60 °C as shown in Figure 3 is more unstable
than in the analogs cells cycled at 25 °C (Figure 2a,b). This indi-
cates that the SEI fracturing at high temperature leads to fur-
ther interfacial reactions at the Cu surface that cause the evolu-
tion of CO2, CO, and C2H4 reductive gases via Schemes 1 and 2.
Therefore, it is challenging to decouple interfacial reactions in
the AFLMBs at high temperature, especially after the first cycle.
Figure S6, Supporting Information, shows the surface mor-
phologies of Cu current collector of Cu/NMC111 full-cells
with EL1 or EL2 cycled at 25, 40, or 60 °C after the third cycle
fully discharge. Some noticeable SEI fracture, which resulted
in repetitive electrolyte contact with the Cu surface, leads to
further electrolyte reactions with Li-metal at high tempera-
ture. For a clear understanding, the XPS measurements of
F-1s, O-1s, and Li-1s spectra were performed for Cu surface of
Cu/NMC111 full cells with EL1 or EL2 electrolyte cycled at 25 or
60 °C as shown in Figures S7 and S8, Supporting Information.
The peaks at 684.5, 685.6, 687, and 688.5 eV in F-1s spectra can

be assigned to LiF, LixPOyFz, C F, and LixPFy, respectively, all
derived from LiPF6 salt decomposition.[37] However, the peaks
at 531.1, 531.8, and 532.8 eV in the O-1s spectra can be attrib-

uted to LiOH, Li2CO3, and C O respectively. The peak at
[25]

535.4 eV in O-1s is assigned to C O. In the Li-1s spectra,
peaks at 53.8, 55.2, and 55.7 eV are assigned to LiO2, Li2CO3,
Adv. Funct. Mater. 2020, 2006951
2006951 (6 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
Figure 4. The reverse LSV of Li/Li/Cu cells prepared with a) EL1, b) EL2 in a potential range of 2–0 V at 0.5 mV s−1 scan rate. Lower panels: reverse
LSV. Upper panels: evolved gases in the cell headspace. The chronoamperometric SEI formation and Li-metal nucleation and growth using Li/Li/Cu
cell with c) EL1, d) EL2. Upper panels: i−t transients of SEI formation at 0.8 V, and nucleation of Li metal deposition on Cu at −0.3 V. Lower panels:
evolved gases in the cell headspace. Cu substrate was used as WE, and Li as CE and RE. The OEMS data were smoothed.
and LiF, respectively while peaks at 54.5 and 56.6 eV attributed
to ROCO2Li and Li2O2 indicate that the SEI is mainly derived
from the solvent decomposition.[38]
evidenced that the evolution of CO2, CO, and C2H4 gases at
the initial stage of the first charging of Cu/NMC111 full-cells is
originated from the interfacial reactions at the Cu surface due
to the electrochemical electrolyte reduction during SEI forma-
tion. The evolved gases in Li/Li/Cu cells cannot affect by the
cathode interfacial reactions since there is no cathode active
material and the voltage range used (2–0 V) is low. Schwenke
et. al.[31] pointed out that EC reduction is characterized by CO2
evolution at 0.8 V. Similarly, the sharp downward peaks starting
at 0.8 V in the LSV curves for both EL1 and EL2 electrolytes is
due to the electrochemical reduction of EC solvent during SEI
formation at Cu surface via Scheme 1, leading to generate CO2,
CO, and C2H4 reductive gases. However, the broad peaks at 1.6
and 1.7 V for the two electrolytes are associated with the elec-
trochemical reduction of DEC and EMC solvent, respectively.
Therefore, the evolution of CO2 and C2H4 gases at 1.6 V in
2.3. Gas Formation Analysis during Electrochemical Reduction
of EL1 and EL2 Electrolytes
Following the study of CO2, CO, C2H4, and C2H6 reductive gas
evolutions at the beginning of the first charging, the poten-
tial-resolved OEMS analysis for Li/Li/Cu cells in the lower
potential range of 2–0 V was conducted. The lower panels in
Figure 4a,b show the potential profiles for cells with EL1 and
EL2 electrolytes, respectively while the upper panels depict the
major evolved gases during reverse linear sweep voltammetry
(LSV) analysis at Cu surface at 0.5 mV s−1 scan rate. The results
Adv. Funct. Mater. 2020, 2006951
2006951 (7 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
Figure 4a and CO2, CO, and C2H4 gases at 1.7 V in Figure 4b is
due to the reduction of DEC and EMC solvents, respectively via
Scheme 2. Since solvents undergo an electrochemical reduction
in the order of EC > EMC > DEC with EC being the most likely
reduced,[24] the current peaks for EC reduction are sharper than
DEC and EMS solvents. Sharp LSV peaks and the intensive
evolution of CO2, CO, and C2H4 gases near to 0.8 V for EL1
and EL2 electrolytes indicates that the SEI formation at Cu is
mainly from EC reduction.
2.4. Gas Formation Analysis of Li/NMC111 Half-Cells at 25 °C
Due to the instability of conventional carbonate electrolytes
with Li anode, the surface reactions in Li/NMC111 half-cells
are different from those in Cu/NMC111 full-cells. The Li anode
chemically react with electrolyte solvents during assembling
and at OCV stage of Li-metal (Li/NMC111) cells, leading to
SEI formation on Li anode. Figure 5a,b show the evolution of
gases associated with the interfacial reactions in Li/NMC111
half-cells with EL1 and EL2 electrolytes, respectively cycled
at 25 °C. For both electrolytes, no gas evolution is observed
before the O2 releasing near to 4.2 V for the cell with EL1 and
4.21 V with EL2 electrolytes. This evidenced that the film
formed at the OCV stage protects electrolyte from electro-
chemical reduction decomposition for SEI formation at the
initial stage of charging. However, the chemically formed SEI
is not stable enough with Li anode and easily fractured at high
voltage. This allows further electrolyte reduction decomposi-
tion at Li anode, leading to the evolution of CO2, CO, C2H4,
C2H6, and POF3 gases.
It is reported that Li2EDC is unstable with Li anode and
easily decompose due to exothermic reactions and SEI expan-
sion in the Li anode in Li/NMC111 cells.[40] Therefore, the evo-
lution of O2 in Li/NMC111 near to 4.2 or 4.21 V for EL1 and EL2
electrolytes, respectively is assigned from the decomposition of
the Li2EDC SEI component on Li anode as shown in Scheme 1.
As a result, the evolution of CO2 and CO gases simultaneously
with O2 releasing in Li/NMC111 is due to the chemical electro-
lyte oxidation via Equation (3). However, the evolution of C2H4
and C2H6 gases at the same time with CO2 and CO evidenced
that there is an electrolyte reaction with Li-metal at the Li anode.
In Cu/NMC11 cells with EL1 electrolyte, the CO2 and CO gases
evolve from the interfacial reactions at the cathode surface near
to 4.38 V in the first cycle. Therefore, in Li/NMC111 half-cells,
The OEMS gas formation analysis was also performed
during SEI formation and Li-metal nucleation and growth
on the Cu current collector by adopting the Thirumalraj et al.
protocol.[39] Figure 4c,d shows the pretreatment (SEI forma-
tion) on Cu current collector at a constant positive potential of
0.8 V for 30 min followed by Li-metal nucleation and growth
at a negative potential of −0.3 V for the next 30 min with a
simultaneous OEMS monitoring of evolved gases. The initial
evolution of CO2, CO and C2H4 gases during the SEI forma-
tion stage indicates that the initial interfacial reaction at the
Cu current collector is mainly due to the electrochemical
reduction of EC solvent. This evidenced that the SEI forma-
tion interfacial reactions are the main source than Li-metal
nucleation and growth for initial gas evolution. The high
intensity of CO gas in the cell with EL2 (Figure 4d) is the char-
acteristic of EMC solvent chemical and/or electrochemical
reduction on Li-metal anode. Thirumalraj et al.[39] have sug-
gested that the SEI is easily fractured during Li-metal growth
on the Cu surface, leading to the electrolyte leakage to the Li
surface. Therefore, the evolution of CO2 gas during Li-metal
nucleation and growth in Li/Li/Cu cells is due to the reaction
between solvents and Li-metal during SEI repairing. No evo-
lution of C2H6 is observed, suggesting that the C2H6 radical
may not form during electrolyte decomposition at a lower
voltage (0−2 V).
Figure 5. Variation of cell voltage (upper panel) and evolved gases (lower panel) versus time profile of Li/NMC111 half-cells with 250 µL of a) EL1,
b) EL2 electrolytes during cycling at 1 mA cm−2 and 25 °C within 2.5–4.8 V voltage range. The OEMS data were smoothed.
Adv. Funct. Mater. 2020, 2006951
2006951 (8 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
the sharp evolution of both CO2 and CO gases starting from 4.2
or 4.21 V up to the end of first charging suggesting that the CO2
and CO gases also evolve from the cathode side due to chemical
and/or electrochemical interfacial electrolyte relations above
4.38 V. As a result, the interfacial reactions could occur at the Li
anode and NMC111 cathode together, conveyed by the evolution
of oxidative gases (CO2, CO, and O2) and reductive gases (C2H4
and C2H6) at the same time. This makes it difficult to decouple
the interfacial reactions at the anode and cathode surface and
more complicated to demystify the initial interfacial reactions
in the Li/NMC111 half-cells. In the second and third cycles
charging of Li/NMC111 half-cells, the evolution of CO2 and CO
gases is due to the electrochemical oxidative decomposition of
electrolyte at the cathode (NMC111) surface. However, the evo-
lution of CO2 and CO gases in the second cycle discharging
(Li stripping) stage is associated with the electrolyte decomposi-
tion at the anode side due to SEI shrinkage and fracturing.
Figure 6a shows the schematic pathways of gas evolution
via chemical and electrochemical interfacial reactions at the
anode and cathode sides in Cu/NMC111 and Li/NMC111 cells.
The Scheme is established based on the OEMS, XPS, and scan-
ning electron microscopy (SEM) data. In the AFLMB, the evo-
lution of gases at separate stages (at the initial stage and high
voltage stage) indicates the decoupling of interfacial reactions
take place at the anode and cathode electrodes. As explained
above, the evolution of CO2, CO, and C2H4 at the initial stage
of first charging of Cu/NMC111 cells indicates the initial inter-
facial reactions at the Cu surface while the evolution of gases
at high voltage (>4.38 V) is originated from chemical and/or
electrochemical interfacial reactions at the cathode surface. In
contrast, in LMBs, the electrolyte chemically reacts with the Li
anode at OCV stage leading to the evolution of CO2, CO, and
C2H4 gases. The SEI film is formed chemically at OCV and pro-
tects the electrolyte from decomposition and no gas evolution
was detected at the initial stage of cell cycling. However, since
it is not compatible with Li anode, the SEI becomes expanding
and fractured above 4.2 V. Therefore, the evolution of CO2, CO,
C2H4, C2H6, and POF3 gases at near to 4.2 is due to the electro-
chemical electrolyte reduction at Li anode during SEI growth.
However, CO2, CO, and O2 gases are occurred from NMC111
side above 4.38 V due to electrolyte oxidation reactions and cou-
pled with gases come from the anode side. Figure 6b,c shows
the rate of gas evolution at OCV stage (the data to the left of
the dashed vertical line) followed by at first cycle charging of
Cu/NMC111 and Li/NMC111 cells, respectively both cells con-
taining EL1 electrolyte. The intensity of gases is too high at the
beginning of gas evolution test at the OCV stage. The reason
why the signal of gases so high at the beginning in Figure 6b,c
is because cells were not purged before starting gas evolution
measurement and the baseline of signals was not subtracted
because our intention was just to study the effect of chemical
reactions at OCV stage of cells. Figure 6b,c are derived plots
from accumulated data in Figures 2a and 5a, respectively.
Therefore, when the baseline is subtracted to remove the data
at OCV stage, Figure 6b,c will be the same as the first cycle
charging of Figure 5a,b, respectively. The rate of evolution of
CO2, CO, and C2H4 gases decreases rapidly from the beginning
of measurement at OCV for the first 3 h (the data to the left side
of the vertical line) in Figure 6b, suggesting that bare Cu was
not chemically reactive with electrolyte and a protective film at
OCV was not formed. However, the CO2, CO, and C2H4 gases
were generated at the beginning of charging of Cu/NMC111
cell (the data right side of the vertical line) associated with the
electrolyte reduction on the Cu surface due to SEI formation.
In contrast, the increasing trend of the rate of CO2, CO, and
C2H4 gases was detected from the beginning of measurement
at the OCV stage of Li/NMC111 cell, as shown in Figure 6c (the
data to the left of the dashed vertical line) evidences the chem-
ical reaction between the electrolyte and Li anode. Once the
Li/MNC111 cell was charged, no gas was detected for the first
1.8 h, suggesting that protective film was formed on Li anode
at OCV stage and protected electrolyte from further decomposi-
tion. Consequently, we can conclude that the evolution of gases
near to 4.2 or 4.1 V is associated with interfacial reactions at the
Cu and NMC111 surfaces.
In Figure 6d, the evolved gases before and after the releasing
of O2 gas upon first cycle charging of Cu/NMC111 and
Li/NMC111 cells discussed in Figures 2 and 5, respectively
were compared. For Cu/NMC111 full-cells with EL1 and EL2
electrolytes, significant amounts of CO2, CO, and C2H4 gases
were evolved before O2 evolution (at <4.39 V for EL1 and
<4.38 V for EL2). However, a high amount of gases are evolved
simultaneously with O2 releasing due to the chemical and/or
electrochemical interfacial reactions at the cathode surface.
The amount of C2H4 reductive gas is equal before and after O2
releasing, indicating that no electrolyte reduction once SEI is
formed. In Li/NMC111 cells, the evolution of gases before O2
releasing (at <4.2 V for EL1 and <4.21 V for EL2) is very insig-
nificant. However, CO2, CO, O2, C2H4, C2H6, and POF3 gases
evolving from the anode and the cathode side above 4.2 V are
mixed.
3. Conclusion
The anode (Li) and cathode (NMC111) interfacial reactions
have been successfully decoupled using a protocol combining
online gas evolution analysis with anode-free Li-metal batteries
(Cu/NMC111) for the first time. The interfacial reactions take
place at both electrodes have been decoupled and compared
in Cu/NMC111 and Li-metal batteries (Li/NMC111) with LiPF6
in ethylene carbonate (EC)/diethyl carbonate (DEC) (1:1 v/v%)
or ethylene carbonate (EC)/ethyl methyl carbonate (EMC)
(1:1 v/v%). The Li anode is chemically reactive with electro-
lytes and forms SEI film at the OCV stage, which results in no
gas evolution in the initial stage of electrochemical cycling of
Li/NMC111 half-cells. However, the SEI film is fractured and
the gaseous by-products such as CO2, CO, O2, C2H4, C2H6, and
POF3 are evolved due to interfacial reactions at Li anode near to
4.2 or 4.21 V for cells with EL1 or EL2 electrolyte, respectively.
The CO2, CO, O2, and POF3 gases also evolved at the cathode
(NMC111) near to 4.38 V and mixed with those formed at Li
anode. This makes it difficult to decouple and demystify the
interfacial reactions in Li/NMC111 cells. However, the inter-
facial reactions at anode and cathode in the Cu/NMC111 full-
cells are successfully decoupled; and noted that the evolution
of CO2, CO, and C2H4 gases at the beginning of first charging
indicates the initial interfacial reactions at Cu surface due
Adv. Funct. Mater. 2020, 2006951
2006951 (9 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
Figure 6. Evolved gases at OCV and during cell cycling were shown at the corresponding stages. a) the schematic representation of gas evolution at
the initial stage and high voltage in the anode-free (AFLMB) and (LMB). b,c) the rate of evolved gases at OCV followed by at first charging of cells.
d) the comparison of evolved gases during the first charging of cells shown in Figures 2 and 5.
to electrochemical electrolyte reduction during SEI forma-
tion. However, the evolution of CO2 and CO gases simultane-
ously with O2 releasing at 4.39 V for EL1 and 4.38 V for EL2
electrolytes in the first cycle as well as in the second and third
cycles is due to both chemical and/or electrochemical interfa-
cial reactions at the cathode surface. At the cell temperature of
Adv. Funct. Mater. 2020, 2006951
2006951 (10 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
namely I28 (CO) = I28 − 1/0.58 * I26. However, since the evolution trends
of CO2 and CO gases were similar, the CO2 fragmentation to yield
CO was insignificant. So the effect of CO ion signal caused by CO2
fragmentation was not considered.
40 or 60 °C, the SEI forming reaction kinetic at the Cu surface
is fast, leading to a high amount of gas evolution at the initial
stage of first charging. The developed protocol and proposed
reaction mechanisms open a new way to decouple interfacial
reactions at both electrodes and help full for developing chemi-
cally and electrochemically stable electrolytes for high perfor-
mance and long-cycling life of AFLMBs and LMBs.
SEI Formation and Li-Metal Nucleation and Growth Protocol for Gas
Analysis: The Li/Li/Cu cells were assembled with EL1 or EL2 electrolyte
using the EL-CELL in the glovebox (where H2O and O2 < 0.1 ppm)
for the analysis of evolved gases during SEI formation and Li-metal
nucleation and growth on bare Cu current collector. The onset potentials
for DEC and EC solvents reduction were first identified from Li/Li/Cu cell
by performing a reverse LSV scan at 0.5 mV s−1 rate in a potential range
of 2–0 V versus Li/Li+. Two potential step protocol was applied for SEI
formation and Li-metal nucleation and growth. The first step was SEI
formation on Cu substrate at a fixed positive potential of 0.8 V versus
Li/Li+ until the current reaches nearly to 0 µA (for 30 min), followed by Li
plating for 30 min using the selected negative applied potential of −0.3 V
versus Li/Li+ in the second step. The potential-controlled electrochemical
experiments were performed using an Autolab potentiostat. All
potentials hereafter are referred to Li/Li+, unless otherwise stated.
Surface Characterization: For Cu surface characterization of
Cu/NMC111 cells with EL1 and EL2 electrolytes cycled at 25 or 60 °C,
the cells were disassembled in argon-filled glovebox after the third cycle
fully discharge. The Cu electrodes cycled with EL1 were washed with DEC
solvent and those cycled with EL2 washed using EMC solvent until all
the residuals were removed and then dried well for surface analysis. The
morphologies of the Cu surface were characterized by SEM (JSM-6500F)
carried out at an accelerating voltage of 15 kV. The compositions of the
SEI on the Cu surface were examined using XPS workstation at beamline
24A1 of the National Synchrotron Radiation Research Center (NSRRC) in
Hsinchu, Taiwan. The XPS was performed using 0.05 eV step and 20 eV
pass energy and calibrated by gold at the binding energy of 84.0 eV.
4. Experimental Section
Electrodes Preparation: Copper foil current collector (Cu) was cut into
15 mm disc and pretreated by washing with 1 m HCl solution for 10 min,
rinsed with deionized water followed by acetone each thrice, then under
vacuum dried for 15 min and transferred into the argon-filled glove
box. LiNi1/3Co1/3Mn1/3O2 (NMC111) cathode material with areal capacity
≈2 mAh cm−2 and 88.5% active material coated on Al-foil was received
from Amita Technologies Inc. (Taiwan). Then, scratch one side by
N-methyl-2-pyrrolidone (99%, Aldrich) solvent and cellulose sponge,
cut into 13 mm diameter disc, followed by vacuum dried at 80 °C
overnight and transferred into glovebox for assembling Cu/NMC111
cells. Li-metal foil (15 mm diameter) was used for assembling Li/
NMC111 cells.
Electrolytes: Commercially available electrolytes, EL1 (1 m LiPF6 in EC/
DEC (1:1 v/v%) and EL2 (1 m LiPF6 in EC/EMC (1:1 v/v%) purchased
from Sigma-Aldrich, and laboratory prepared 1 m LiPF6 in EC were used
without further purification. In the following sections, the electrolytes
would be described as written to minimize the redundancy of common
terms, symbols, and units.
OEMS Gas Formation Analysis: The homemade EL-CELL of headspace
4 mL containing Cu/NMC111, Li/NMC111, or Li/Li/Cu cell with 250 µL
EL1 or EL2 using Celgard 2325 separator were assembled in the argon-
filled glove box (where H2O and O2 <0.1 ppm) in the order shown in
Figure S1, Supporting Information. The glass fiber separator was used
for the 1 m LiPF6 in EC solvent electrolyte assembled in Cu/NMC111 cell.
The assembled EL-CELL was placed into the temperature-controlled
chamber (BMNS-500, Heating Mantle, Newlab. Instruments Co., Ltd,
Germany) and connected to GC-MS (QP2010 Ultra SHIMADZU) using
ethylene tetrafluoroethylene tube (ETFE, 45 cm length and 1.0 mm
inner diameter) as illustrated in Figure 1. The cell then held at an
open-circuit voltage (OCV) at 25 °C and purged with a helium carrier
gas (99.9995% purity) at an optimized purge flow rate of 1 mL min−1
for 2 h to remove any residual trace gases from the cells. All gases
such as CO2 (m/z = 44), CO (m/z = 28), O2 (m/z = 32), H2 (m/z = 2),
C2H4, (m/z = 26), C2H6 (m/z = 30) and POF3 (m/z = 85) have been
then measured at OCV for 3 h in order to obtain stable m/z signal
background for the proper baseline subtraction.[41] The total and column
flow rates were 14.9 and 0.45 mL min−1, respectively. All GC-MS protocols
are listed in Table S1, Supporting Information. After 3 h gas evolution
measurement at OCV stage, cells were cycled using constant-current,
constant-voltage/constant-current (CCCV/CC) mode by using computer-
controlled potentiostat/galvanostat (PGSTAT302N, Metrohm Autolab)
at 1 mA cm−2 current density within 2.5–4.8 V versus Li/Li+ voltage
range without stopping OCV stage OEMS gas evolution measurement.
The evolved gases (CO2 (m/z = 44), CO (m/z = 28), O2 (m/z = 32), H2
(m/z = 2), C2H4, (m/z = 26), C2H6 (m/z = 30) and POF3 (m/z = 85)) were
measured continuously during charge/discharging of cells for the first
three cycles. All the OEMS data were smoothed by 200 points to avoid
signal noises.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
Acknowledgements
Financial support from the Ministry of Science and Technology of
Taiwan (108-3116-F-011-001-CC1, 108-2627-M-011-001-, 107-2923-E-011-
002, 107-2119-M-002-033-, 106-2923-E 011-005, 106-2221-E-011-125-MY3),
the Ministry of Education of Taiwan (U2RSC program, MOE 1080059),
Taiwan’s Deep Decarbonization Pathways toward a Sustainable Society
Project (AS-KPQ-106- DDPP) from Academia Sinica as well as the
facilities of support from National Taiwan University of Science and
Technology (NTUST) and National Synchrotron Radiation Research
Centre (NSRRC) are all gratefully acknowledged.
Conflict of Interest
The authors declare no conflict of interest.
Author Contributions
All cells were cycled at a temperature of 25, 40, or 60 °C with 1 °C s−1
flow rate of temperature from ambient temperature to fixed one. The
K.N.S. developed the methodology, investigated and performed
experiments, characterized and analyzed the results, and wrote the
manuscript. S.-C.Y. developed GC-MS methodology, electrochemical
characterization, and data analysis. S.-K.J. and C.-J.H. coordinated
all experimental works and contributed to GC-MS analysis, SEM,
and XPS measurements. N.A.S., Y.N., and H.H.W. contributed to the
technical activities and experimental works. C.-H.W facilitated the XPS
fragment of C2H4 with m/z = 26 was used to measure C2H4 to avoid the
[8,9]
interference of CO (m/z = 28) ion signal to the ion signal of C2H4
.
From the GC-MS library database, the ion signal of m/z = 26 from C2H4
fragmentation was 58% of C2H4 (m/z = 28) signal. So, the actual ion
signal of CO (I28) was obtained by subtracting the C2H4 ion signal (I26)
multiplying a correction factor 1/(0.58) from CO ion signal (I28),[41,42]
Adv. Funct. Mater. 2020, 2006951
2006951 (11 of 12)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.afm-journal.de
measurement, S.-H.W. contributed to the conceptualization, visualization,
writing the review, and editing the manuscript. W.-N.S. contributed to the
conceptualization, funding acquisition, visualization, writing the review,
and editing the manuscript. B.J.H. contributed to the conceptualization,
supervision, funding acquisition, project administration, resources,
visualization, writing the review, and editing the manuscript. All the
authors discussed the results and commented on the manuscript.
[19] D. M. Seo, D. Chalasani, B. S. Parimalam, R. Kadam, M. Nie,
B. L. Lucht, ECS Electrochem. Lett. 2014, 3, A91.
[20] S. D. Xu, Q. C. Zhuang, J. Wang, Y. Q. Xu, Y. B. Zhu, Int. J. Electro-
chem. Sci. 2013, 8, 8058.
[21] D. Pritzl, S. Solchenbach, M. Wetjen, H. A. Gasteiger, J. Electro-
chem. Soc. 2017, 164, A2625.
[22] a) A. Wang, S. Kadam, H. Li, S. Shi, Y. Qi, npj Comput. Mater. 2018,
4, 15; b) G. M. Hobold, A. Khurram, B. M. Gallant, Chem. Mater.
2020, 32, 2341; c) J. Young, M. Smeu, J. Phys. Chem. Lett. 2018, 9,
3295; d) G. Gachot, S. Grugeon, M. Armand, S. Pilard, P. Guenot,
J.-M. Tarascon, S. Laruelle, J. Power Sources 2008, 178, 409.
[23] S. Solchenbach, M. Metzger, M. Egawa, H. Beyer, H. A. Gasteiger,
J. Electrochem. Soc. 2018, 165, A3022.
Keywords
anode-free batteries, gas evolution, interfacial reactions, Li-metal
batteries, online electrochemical mass spectroscopy
[24] K. Tasaki, J. Phys. Chem. B 2005, 109, 2920.
[25] S. J. An, J. Li, C. Daniel, D. Mohanty, S. Nagpure, D. L. Wood,
Carbon 2016, 105, 52.
[26] T. Nakajima, Y. Hirobayashi, Y. Takayanagi, Y. Ohzawa, J. Power
Sources 2013, 243, 581.
Received: August 17, 2020
Revised: September 30, 2020
Published online:
[27] T. Li, X.-Z. Yuan, L. Zhang, D. Song, K. Shi, C. Bock, Electrochem.
Energy Rev. 2020, 3, 43.
[1] a) H. Ye, S. Xin, Y.-X. Yin, J.-Y. Li, Y.-G. Guo, L.-J. Wan, J. Am. Chem.
Soc. 2017, 139, 5916; b) Y. Yang, D. M. Davies, Y. Yin, O. Borodin,
J. Z. Lee, C. Fang, M. Olguin, Y. Zhang, E. S. Sablina, X. Wang,
C. S. Rustomji, Y. S. Meng, Joule 2019, 3, 1986; c) W. Xu, J. Wang,
F. Ding, X. Chen, E. Nasybulin, Y. Zhang, J.-G. Zhang, Energy
Environ. Sci. 2014, 7, 513.
[2] a) J. Qian, B. D. Adams, J. Zheng, W. Xu, W. A. Henderson, J. Wang,
M. E. Bowden, S. Xu, J. Hu, J.-G. Zhang, Adv. Funct. Mater. 2016, 26,
7094; b) J.-J. Woo, V. A. Maroni, G. Liu, J. T. Vaughey, D. J. Gosztola,
K. Amine, Z. Zhang, J. Electrochem. Soc. 2014, 161, A827; c) D. Lin,
Y. Liu, Y. Cui, Nat. Nanotechnol. 2017, 12, 194.
[3] M. S. Milien, H. Beyer, W. Beichel, P. Klose, H. A. Gasteiger,
B. L. Lucht, I. Krossing, J. Electrochem. Soc. 2018, 165, A2569.
[4] a) X.-B. Cheng, R. Zhang, C.-Z. Zhao, Q. Zhang, Chem. Rev. 2017,
117, 10403; b) L. Li, S. Li, Y. Lu, Chem. Commun. 2018, 54, 6648;
c) G. Zheng, Y. Xiang, S. Chen, S. Ganapathy, T. W. Verhallen, M. Liu,
G. Zhong, J. Zhu, X. Han, W. Wang, W. Zhao, M. Wagemaker,
Y. Yang, Energy Storage Mater. 2019.
[5] C. P. Aiken, J. Xia, D. Y. Wang, D. A. Stevens, S. Trussler, J. R. Dahn,
J. Electrochem. Soc. 2014, 161, A1548.
[6] J. Self, C. P. Aiken, R. Petibon, J. R. Dahn, J. Electrochem. Soc. 2015,
162, A796.
[7] R. Jung, P. Strobl, F. Maglia, C. Stinner, H. A. Gasteiger, J. Electro-
chem. Soc. 2018, 165, A2869.
[8] M. Metzger, B. Strehle, S. Solchenbach, H. A. Gasteiger, J. Electro-
chem. Soc. 2016, 163, A798.
[28] a) S. A. Freunberger, Y. Chen, Z. Peng, J. M. Griffin, L. J. Hardwick,
F. Bardé, P. Novák, P. G. Bruce, J. Am. Chem. Soc. 2011, 133, 8040;
b) N. Yabuuchi, K. Yoshii, S.-T. Myung, I. Nakai, S. Komaba, J. Am.
Chem. Soc. 2011, 133, 4404; c) M. Balaish, A. Kraytsberg, Y. Ein-Eli,
Phys. Chem. Chem. Phys. 2014, 16, 2801; d) F. Lian, Y. Li, Y. He,
H. Guan, K. Yan, W. Qiu, K.-C. Chou, P. Axmann, M. Wohlfahrt-
Mehrens, RSC Adv. 2015, 5, 86763.
[29] a) J. Wandt, A. T. S. Freiberg, A. Ogrodnik, H. A. Gasteiger, Mater.
Today 2018, 21, 825; b) A. T. S. Freiberg, M. K. Roos, J. Wandt,
R. de Vivie-Riedle, H. A. Gasteiger, J. Phys. Chem. A 2018, 122, 8828.
[30] L. Xing, W. Li, C. Wang, F. Gu, M. Xu, C. Tan, J. Yi, J. Phys. Chem. B
2009, 113, 16596.
[31] K. U. Schwenke, S. Solchenbach, J. Demeaux, B. L. Lucht,
H. A. Gasteiger, J. Electrochem. Soc. 2019, 166, A2035.
[32] S. Wilken, M. Treskow, J. Scheers, P. Johansson, P. Jacobsson, RSC
Adv. 2013, 3, 16359.
[33] C. Schultz, S. Vedder, B. Streipert, M. Winter, S. Nowak, RSC Adv.
2017, 7, 27853.
[34] J. Lamb, C. J. Orendorff, E. P. Roth, J. Langendorf, J. Electrochem.
Soc. 2015, 162, A2131.
[35] K. Leung, F. Soto, K. Hankins, P. B. Balbuena, K. L. Harrison,
J. Phys. Chem. C 2016, 120, 6302.
[36] M. Onuki, S. Kinoshita, Y. Sakata, M. Yanagidate, Y. Otake, M. Ue,
M. Deguchi, J. Electrochem. Soc. 2008, 155, A794.
[37] a) X.-Q. Zhang, X.-B. Cheng, X. Chen, C. Yan, Q. Zhang, Adv. Funct.
Mater. 2017, 27, 1605989; b) L. Liu, S. Gu, S. Wang, X. Zhang,
S. Chen, RSC Adv. 2020, 10, 1704.
[9] N. E. Galushkin, N. N. Yazvinskaya, D. N. Galushkin, J. Electrochem.
Soc. 2019, 166, A897.
[10] W. Li, B. Song, A. Manthiram, Chem. Soc. Rev. 2017, 46, 3006.
[11] A. Guéguen, C. Bolli, M. A. Mendez, E. J. Berg, ACS Appl. Energy
Mater. 2020, 3, 290.
[12] K. Xu, Chem. Rev. 2004, 104, 4303.
[13] T. Yoon, T. Lee, J. Soon, H. Jeong, S. Jurng, J. H. Ryu, S. M. Oh,
J. Electrochem. Soc. 2018, 165, A1095.
[38] T. T. Hagos, B. Thirumalraj, C.-J. Huang, L. H. Abrha, T. M. Hagos,
G. B. Berhe, H. K. Bezabh, J. Cherng, S.-F. Chiu, W.-N. Su,
B.-J. Hwang, ACS Appl. Mater. Interfaces 2019, 11, 9955.
[39] B. Thirumalraj, T. T. Hagos, C.-J. Huang, M. A. Teshager,
J.-H. Cheng, W.-N. Su, B.-J. Hwang, J. Am. Chem. Soc. 2019, 141,
18612.
[14] a) S. S. Zhang, Front. Energy Res. 2014, 2, 59; b) A. Kriston, I. Adanouj,
V. Ruiz, A. Pfrang, J. Power Sources 2019, 435, 226774; c) R. Jung,
M. Metzger, F. Maglia, C. Stinner, H. A. Gasteiger, J. Phys. Chem.
Lett. 2017, 8, 4820.
[40] a) G. Ramos-Sanchez, F. A. Soto, J. M. Martinez de la Hoz, Z. Liu,
P. P. Mukherjee, F. El-Mellouhi, J. M. Seminario, P. B. Balbuena,
J. Electrochem. Energy Convers. Storage 2016, 13, 031002;
b) R. Spotnitz, J. Franklin, J. Power Sources 2003, 113, 81; c) T. Liu,
L. Lin, X. Bi, L. Tian, K. Yang, J. Liu, M. Li, Z. Chen, J. Lu, K. Amine,
K. Xu, F. Pan, Nat. Nanotechnol. 2019, 14, 50; d) J. Henschel,
C. Peschel, S. Klein, F. Horsthemke, M. Winter, S. Nowak, Angew.
Chem., Int. Ed. 2020, 59, 6128.
[15] Y. Li, Y. Qi, Energy Environ. Sci. 2019, 12, 1286.
[16] Y. Jin, N.-J. H. Kneusels, P. C. M. M. Magusin, G. Kim, E. Castillo-
Martínez, L. E. Marbella, R. N. Kerber, D. J. Howe, S. Paul, T. Liu,
C. P. Grey, J. Am. Chem. Soc. 2017, 139, 14992.
[17] M. He, L. Boulet-Roblin, P. Borel, C. Tessier, P. Novák, C. Villevieille,
E. J. Berg, J. Electrochem. Soc. 2016, 163, A83.
[41] M. Metzger, C. Marino, J. Sicklinger, D. Haering, H. A. Gasteiger,
J. Electrochem. Soc. 2015, 162, A1123.
[18] Y. Bi, T. Wang, M. Liu, R. Du, W. Yang, Z. Liu, Z. Peng, Y. Liu,
D. Wang, X. Sun, RSC Adv 2016, 6, 19233.
[42] B. Strehle, S. Solchenbach, M. Metzger, K. U. Schwenke,
H. A. Gasteiger, J. Electrochem. Soc. 2017, 164, A2513.
Adv. Funct. Mater. 2020, 2006951
2006951 (12 of 12)
© 2020 Wiley-VCH GmbH