Journal of Energy Chemistry 60 (2021) 9–15

Contents lists available at ScienceDirect

Journal of Energy Chemistry

journal homepage: www.elsevier.com/locate/jechem

Real-time monitoring of the lithiation process in organic electrode

7,7,8,8-tetracyanoquinodimethane by in situ EPR
Mingxue Tang a,b, Nhat N. Bui b, Jin Zheng a,b, Likai Song b, Yan-Yan Hu a,b,⇑
a
Department of Chemistry and Biochemistry, Florida State University, Tallahassee, FL 32306, USA
b
National High Magnetic Field Laboratory, 1800 East Paul Dirac Drive, Tallahassee, FL 32310, USA

a r t i c l e i n f o a b s t r a c t

Article history: Organic electrodes are advantageous for rechargeable lithium-ion batteries owing to their high theoret-
Received 11 November 2020 ical capacities, diverse functionalities, and environmental compatibility. Understanding the working
Revised 11 December 2020 mechanism of organic electrodes is vital to strategic materials design. However, due to lack of suitable
Accepted 15 December 2020
characterization tools, it has been challenging to probe the reaction processes of organic electrodes in
Available online 31 December 2020
real-time. Here, non-destructive in situ electron paramagnetic resonance (EPR) was performed on a model
organic electrode, 7,7,8,8-tetracyanoquinodimethane (TCNQ) used in rechargeable lithium-ion batteries,
Keywords:
to directly follow the redox reactions in real-time. In order to minimize interfering signals from other
Organic electrode
In situ EPR
parts of the batteries than the TCNQ electrode of interest, two sets of batteries are fabricated and studied
Rechargeable batteries with in situ EPR: (1) a LiCoO2//Li4Ti5O12 full-cell battery to determine the EPR signal evolution of additives
7,7,8,8-tetracyanoquinodimethane and electrolytes; (2) a LiCoO2//TCNQ battery, and the difference in the observed EPR signals reflects
purely the redox reactions of TCNQ upon lithiation and delithiation. A two-electron reversible redox reac-
tion is delineated for TCNQ. TCNQ dimers form during the first electron injection upon lithiation and fol-
lowed by the break-down of the dimers and associated electron coupling to produce massive delocalized
electrons, resulting in increased EPR signal during the 2nd electron injection. Reversible trends are
observed during electron ejection upon delithiation. In situ EPR is very sensitive to electron activities,
thus is a powerful tool to follow redox reactions of organic electrodes, allowing for improved fundamen-
tal understanding of how organic electrodes work and for informed design of high-performance organic
materials for energy storage.
Ó 2020 Science Press and Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by
ELSEVIER B.V. and Science Press. All rights reserved.

1. Introduction solubility in organic solvents [6,11,12]. Very recently, metal–or-

ganic framework (MOF) has been proposed as another way to
Electrodes with high energy and high power densities, good sta- improve the cycling performance together with increased capacity
bility, low cost, and environmental compatibility are in need for by additional transition metal redox [13,14].
high-performance rechargeable batteries [1–5]. The specific capac- Among all the organic electrodes, quinone-based materials
ities of inorganic electrodes are limited by the redox reactions of [8,12,15] with cyanide groups show high capacity and good cycling
transition metal elements and structural rigidity. By contrast, reversibility [16–20]. However, the lithiation and delithiation
organic electrodes show potential in producing high gravimetric mechanisms of these materials are elusive and fundamental under-
capacity via multi-electron reactions through functionalization. standing of how these organic electrodes work in rechargeable bat-
In addition, their sustainability and environmental compatibility teries in real-time will provide guidelines on how to modify their
are also appealing [6–8]. However, a major challenge for organic structures to achieve improved electrochemical performance.
electrodes is capacity degradation due to decomposition and or 7,7,8,8-tetracyanoquinodimethane (TCNQ) is one of the represen-
solubility in electrolytes. Polymerization and salt formation are tative organic electrodes used in rechargeable lithium-ion batter-
two effective approaches to improve their cycling stability ies. Extensive efforts have been directed towards understanding
[7,9,10]. Solid-state batteries can circumvent the issue of high the interactions of Li+ and e- within the organic structural frame-
work. Lithiation and delithiation are usually accompanied by
changes in electron band structures and charge re-distribution.
⇑ Corresponding author.
Some reports claimed that free radicals are formed upon lithiation
E-mail address: [email protected] (Y.-Y. Hu).

https://doi.org/10.1016/j.jechem.2020.12.009
2095-4956/Ó 2020 Science Press and Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by ELSEVIER B.V. and Science Press. All rights reserved.
M. Tang, N.N. Bui, J. Zheng et al. Journal of Energy Chemistry 60 (2021) 9–15

Fig. 1. The experimental setup for the in situ EPR acquisition and the EPR spectra from control experiments performed on battery components. (a) Schematic of a bag-cell
battery for in situ EPR characterization positioned in the EPR cavity in a magnet as shown in (b). The battery is connected to a Gamry potentiostat and the EPR console for
simultaneous acquisition of electrochemical profiles and EPR spectra. (c) EPR spectra of the components used within the batteries. (d) EPR spectra of an LCO//TCNQ and an
LCO//LTO bag-cell battery acquired right after assembly.

of TCNQ [6], while others based on computational studies pro- delithiation. The gained new knowledge sheds light on the working
posed TCNQ dimer formation [21,22]. However, no direct experi- mechanism of TCNQ electrodes and other quinone-based organic
mental evidence has been made available to either argument. materials in rechargeable Li-ion batteries and provides insights
Various techniques, such as Infrared, Ramman, UV–vis, Nuclear into structural design for improved electrochemical performance
Magnetic Resonance (NMR), and Scanning Transmission Electron of organic electrodes.
Spectroscopy (STEM) have been explored to determine the struc-
ture and working mechanism of organic electrodes 2. Experimental
[9,12,15,17,18,23]. In particular, electron paramagnetic resonance
(EPR), which is a sensitive technique for probing unpaired, conduc- 2.1. Materials
tive, and delocalized electrons, has been proved as a non-
destructive and powerful tool for both ex situ [24–33] and in situ 7,7,8,8-tetracyanoquinodimethane (TCNQ), lithium cobalt oxide
investigations of changes in electron states [34–41]. In this contri- (LiCoO2, 98%), and lithium titanate spinel oxide (Li4Ti5O12, particle
bution, in situ EPR is employed to determine the states and distri- size < 100 nm) were purchased from Sigma Aldrich. They were
bution of electrons in TCNQ electrodes upon lithiation and dried at 120 ℃ for two days prior to making electrodes. Binder
10
M. Tang, N.N. Bui, J. Zheng et al. Journal of Energy Chemistry 60 (2021) 9–15

Fig. 2. In situ EPR spectra of an LCO//LTO bag-cell battery upon the 1st charge/discharge electrochemical cycling. (a) Overlapped and (b) sequentially displayed in situ EPR
spectra for the first cycle and (c) the electrochemical profile and EPR spectral areal integrals of the LCO//LTO bag-cell battery as a function of charge states.

polyvinylidene fluoride (PVDF, Kynar Flex 2801, Arkema) and 60 EPR background is only originated from the evolution of acetylene
wt% solution of polytetrafluoroethylene (PTFE, batch number black, which is carefully subtracted with our experiential design.
#0011601, MTI Corporation) were used as received without further Therefore, the difference in binder and ratio does not affect the
purification. data analysis and final conclusion.
For in situ EPR measurements, plastic bag cells were assembled
2.2. Sample preparation and characterization in an Argon-filled glove-box and were electrochemically cycled
inside an EPR cavity. The cell was cycled with a Gamry 600 + poten-
2.2.1. Electrochemistry tiostat. EPR spectra were collected while cycling the batteries at a
The inorganic electrodes were made by mixing LiCoO2 (LCO) rate of C/15, with the capacity of C = 130 and 260 mAh
g 1 for LCO
and Li4Ti5O12 (LTO) with polyvinylidene fluoride (PVDF) and con- and TCNQ, respectively.
ductive acetylene black (80:10:10, w:w:w). The mixture was dis-
persed in N-Methyl-2-pyrrolidone (NMP) and manually ground 2.2.2. EPR measurements
in an agate mortar for 30 minutes to form a homogenous slurry. Continuous-wave EPR (CW-EPR) spectra were acquired on a
The slurry was cast onto a Ti mesh and dried at 120 °C for 4 h under Bruker Elexsys E680 spectrometer at X-band (9.5 GHz) (Bruker
vacuum before assembled into batteries inside an Argon-filled BioSpin, Billerica, MA), using a Bruker High Sensitivity cavity (ER
glove-box. 4119HS) equipped with a home-made holder to secure the battery.
A similar procedure was employed to make organic TCNQ elec- All spectra during battery cycling were recorded under the kinetics
trodes. The active material TCNQ, binder PTFE, and acetylene black mode (the F2 dimension is set to time). Each spectrum was the
were mixed with a weight ratio of 60:10:30. Ethanol was used as average of 6 scans with a total scanning time of 18 minutes. An
the paste solvent during mixing in order to improve the contact interval of 150 ms was set between the spectra. The recording win-
between TCNQ and acetylene. The film obtained after ethanol dow was set to 200 mT with 4096 data points, covering all the EPR
evaporation was cut into small pieces, pressed onto Ti mesh, and resonances in the current study. The microwave power and modu-
dried under vacuum at 90 °C for 2 h. The obtained electrodes were lation amplitude were set to 0.2 mW and 0.5 mT, respectively. The
transferred into an Argon-filled glove-box for battery assembly. areal integrals of the recorded spectra were calculated using
The LCO cathode and LTO or TCNQ anode were separated by a piece MATLAB and Easyspin [42]. The first data point was normalized
of porous glass microfiber (Whatman, type GF/D), which was to be 1.
soaked in the LP30 electrolyte, i.e., 1 M LiPF6 in ethylene carbonate
and dimethyl carbonate (1:1, w:w, Merck). 3. Results and discussion
It is worth noting that different binders and conductive additive
amount were used for LCO/LTO and LCO/TCNQ based on their elec- To investigate the lithiation and delithiation mechanism of
tronic conductivities and physical properties, which was necessary TCNQ, a careful design of a full battery is necessary to avoid inter-
to make the battery cells work. Nevertheless, the binders are EPR ference of EPR signals from other parts of the battery in addition to
silent and examined with careful control experiments, and the those originating from TCNQ. To prepare the battery cell, EPR
11
M. Tang, N.N. Bui, J. Zheng et al. Journal of Energy Chemistry 60 (2021) 9–15

Fig. 3. In situ EPR of an LCO//TCNQ bag-cell battery upon electrochemical cycling. (a) EPR spectra of the freshly assembled battery at the top of charge and after one cycle,
respectively. (b) All in situ EPR spectra for the first charge–discharge cycle, along with the corresponding electrochemical profile.

experiments are first performed on candidate active electrode Ti3+ is negligible. Therefore, an LCO//LTO full battery is employed
materials and supporting components. Microwave penetration is to determine the EPR signal evolution of acetylene black upon
often limited by metallic components of battery cells; therefore, cycling, which will later be used for background correction to
bag-cell batteries are fabricated with Ti mesh as current collectors obtain the EPR signal purely from TCNQ. Thus, two sets of full bat-
to maximize microwave (MW) penetration and in situ EPR sensitiv- teries (LCO//LTO and LCO//TCNQ) are investigated. Both pristine
ity. Preparatory experiments have shown that the separator (glass LCO//TCNQ and LCO//LTO bag-cell batteries show asymmetric
fiber), air-tight plastic cover for the bag-cells, and Ti current collec- EPR spectra (Fig. 1d), which is possibly caused by organic carbon-
tors are EPR silent at room temperature. For in situ EPR acquisi- ates coordinated to nearby Li+ ions in the LP30 electrolyte [45–48],
tions, the assembled bag-cell battery (Fig. 1a) is vertically resembling the effect of the conductive bulk metals
positioned in the EPR cavity (Fig. 1b) and connected to a Gamry [34,39,36,49,50]. The EPR signal of the LCO//TCNQ battery origi-
potentiostat for electrochemical operation. Control EPR experi- nates from both acetylene black and TCNQ; the EPR signal of the
ments are first carried out on individual components (Fig. 1c) used LCO//LTO battery is from acetylene alone.
in the batteries. Conductive acetylene black shows relatively In order to evaluate the contribution to the EPR signal by acet-
intense EPR signals. Pristine TCNQ displays a weak EPR resonance ylene black used as the conductive additive in the battery, the real-
due to delocalized electrons within its core [43]. The EPR spectrum time evolution of the EPR signal from acetylene black is acquired
of the electrodes containing a mixture of TCNQ, poly tetra fluo- on an LCO//LTO battery. In situ EPR spectra of the LCO//LTO battery
roethylene (PTFE), and acetylene black with a weight ratio of 6: during the 1st charge–discharge are shown in Fig. 2a and 2b. The
1: 3, is also acquired and shown in Fig. 1c (second spectrum from EPR spectra exhibit the same line shape with intensity gradually
the top). Since the EPR resonances of TCNQ and Li metal overlap increases upon charge and decreases upon discharge. The normal-
with each other, to avoid interference for quantitative analysis of ized spectral areal integral is plotted as a function of capacity
EPR for TCNQ, EPR-silent LCO is chosen as the counter electrode. shown in Fig. 2c, along with the electrochemical profile of the
Both pristine LCO and LTO are EPR silent at room temperature at LCO//LTO battery cycled at a rate of C/15 with C = 137 mAh/g cal-
X-band as shown at the bottom of Fig. 1c, most likely due to large culated based on the theoretical capacity of LCO by removing 0.5 Li
zero-field splitting and fast relaxation [44]. In fact, Ti3+ is EPR per formula. Upon charge, about 0.4 Li is extracted from LCO and
active and weak signals could be detected if the measured sample intercalated into LTO; upon discharge, 0.35 Li is deintercalated
is under dry conditions. In this in situ study, the EPR signal from from LTO and reinserted back into LCO. The EPR signal intensity

12
M. Tang, N.N. Bui, J. Zheng et al. Journal of Energy Chemistry 60 (2021) 9–15

Fig. 4. Quantitative analysis of the in situ EPR on an LCO//TCNQ battery. (a) The electrochemical profile of the LCO//TCNQ battery during in situ EPR acquisition. (b) Normalized
EPR spectral areal integral as a function of the specific capacity for the LCO//TCNQ battery. The areal integral of the EPR signal from the pristine battery is normalized to be 1.
For reference, the spectral areal integral of EPR signal originating from components in the battery other than TCNQ is shown as dark brown dots, determined based on the
in situ EPR control experiments on an LCO/LTO battery (Fig. 2).

of acetylene black is increased by about 10% at the top of charge corresponding electrochemical profiles shown below. Only slight
compared with the pristine state, which is probably attributed to increase in the EPR signal intensity of the LCO//TCNQ battery is
the redistribution of conductive electrons during the charge pro- observed at the beginning of lithiation of TCNQ, till reaching a
cess. This change is almost recovered upon discharge. The EPR evo- capacity of 80 mAh/g. The growth of the EPR signal is significantly
lution of acetylene black used in these batteries will be taken into faster for the rest of the charging (lithiation) process. The reverse
account for the following quantitative studies. pattern of signal change is observed upon delithiation of TCNQ.
To help elucidate the working mechanism of TCNQ electrodes in The evolution of the in situ EPR signal reflects changes in redox
rechargeable Li-ion batteries, in situ EPR experiments are per- reactions of TCNQ. Further quantitative analysis of these changes
formed on a bag-cell battery made of LCO//TCNQ. Since LCO is will afford more in-depth insights into the working mechanism
EPR silent at room temperature, it is used as the counter electrode of TCNQ electrodes in situ.
to TCNQ and provides the necessary Li+ source. The pristine LCO// Quantitative analysis of changes in the EPR signal intensity of
TCNQ battery shows an asymmetric EPR resonance (Fig. 3a) mainly an LCO//TCNQ battery cell over the first three lithiation-
originating from the conductive additive acetylene black. The sig- delithiation cycles is performed and the results are summarized
nal from pristine TCNQ contributes a very small fraction to the in Fig. 4b, along with the corresponding electrochemical profile
overall EPR resonance of the full battery, suggested by the results of the LCO//TCNQ bag-cell battery shown in Fig. 4a. About 0.9 Li
from the control experiments shown in Fig. 1c. However, at the is intercalated into TCNQ during the 1st plateau at 0.75 V and 0.3
top of charge (lithiated TCNQ), both positive and negative magni- Li for the 2nd plateau at 1.4 V. The obtained capacities are lower
tudes grow significantly (spectrum in blue in Fig. 3b). The subtrac- than the theoretical value, i.e., 1 Li for each plateau, which is attrib-
tion of the spectrum of the pristine battery from that of the top uted to a few reasons: i) inadequate amount of electrolyte is used
lithiated TCNQ gives rise to the change in EPR signal due to the in the batteries, since –CH3 group within the solvents of LP30 elec-
lithiation process. The difference spectrum is plotted in green trolytes greatly affects EPR cavity tuning and poor tuning results in
and labeled with TCNQ2- at the top of charge, as theoretically it will significantly reduced detection sensitivity; ii) dissolution of TCNQ
form a structure with the composition of Li2TCNQ. The difference in the electrolyte; iii) low electronic conductivity of the TCNQ elec-
spectrum exhibits a symmetric shape, comparable to the ones from trode, yielding a fraction of TCNQ in the electrode not accessible.
control experiments on electrolyte-free samples (Fig. 1c). At the All these factors will lead to capacity reduction, which will be con-
end of the first charge–discharge cycle, the EPR spectrum is nearly sidered in the following analysis. Upon discharge, 0.3 Li is
recovered to the one acquired on the pristine battery (spectrum in extracted from TCNQ over the first plateau at 1.15 V and 0.6 Li dur-
red in Fig. 3a). The whole series of in situ EPR spectra following the ing the 2nd plateau at about 0.5 V. A hysteresis of 0.25 V is
first charge–discharge cycle is plotted in Fig. 3b, along with the observed for both plateaus during the lithiation process of TCNQ.

13
M. Tang, N.N. Bui, J. Zheng et al. Journal of Energy Chemistry 60 (2021) 9–15

Fig. 5. Proposed mechanism for (de)lithiation of TCNQ within an LCO//TCNQ battery, based on this in situ EPR study.

These patterns are observed beyond the first cycle with the same in situ EPR signal rapidly grows in almost a linear fashion, indicat-
cycling rate of C/15 (Fig. 4a). Upon further charge, about 0.7 Li at ing continuous formation of TCNQ monomer radicals. The reversi-
the lower plateau and 0.4 Li at the higher plateau are intercalated ble process is observed upon delithiation. The same trend is
into TCNQ. 0.3 Li at the higher plateau and 0.7 Li at the lower pla- observed but with increased maximum intensity for the next two
teau are extracted from TCNQ upon discharge. The irreversible lithiation-delithiation cycles. This is mainly due to increased inter-
capacity is possibly attributed to the low electronic conductivity calation of Li into TCNQ, which contains 1.2, 1.4, and 1.5 Li to total
of state of change TCNQ and insufficient electrolyte. 1.1 Li is further at top of charge during the three cycles. In addition, the disorder
intercalated into TCNQ and 0.9 Li is extracted upon the 3rd charge caused by cycling can possibly promote delocalization of electrons
and discharge. The 2nd plateau is shorter than the 1st one dis- with enhanced EPR signal as well [24,26,28,51].
played in the three presented cycles is possibly due to the high According to the above results and discussion, the reaction
resistance caused by insufficient electrolyte. Although full capacity mechanism of lithiation and delithiation of TCNQ is proposed as
is not achieved, the following analysis and results can sufficiently shown in Fig. 5. During lithiation, Li+ inserts between two cyanide
support the two-electron reaction mechanism. groups in TCNQ, consistent with previous computational studies
Spectral areal integrals of the in situ EPR of the LCO//TCNQ bat- [22,52]. Meanwhile, e- imparts into TCNQ, resulting in electron
tery mentioned above are plotted in Fig. 4b, together with those of re-distribution (or chemical bond re-configuration) within the
an LCO//LTO battery for reference. The former set of integrals entire molecule and transforming a quinone ring into a benzene
include contributions from TCNQ, acetylene black, and electrolyte, ring. The unpaired electron within each lithiated-TCNQ molecule
and the latter from acetylene black and electrolyte; therefore, the promotes its coupling to another lithiated-TCNQ, forming a pi-
difference between these two sets of integrals reflects evolution stacking dimer. This explains the reason why early-stage lithiation
of EPR signals purely from TCNQ in the electrode. As shown in of TCNQ does not yield net EPR signal gain. Upon further lithiation,
Fig. 4b, at the initial lithiation of TCNQ up to 0.7 Li+/TCNQ, no net the inserted Li+ ions break down the previously formed dimers,
change in the EPR resonance from TCNQ or lithiated-TCNQ is yielding Li+-TCNO-, which allows further delocalization of electrons
observed, suggesting that no radicals are formed or significant within the entire molecule, resulting in increased EPR signal. Upon
change in ERP-relevant electron properties. Near the end of the delithiation, a reversible process occurs, including the dimer for-
1st plateau and during the entire 2nd plateau upon lithiation, the mation at the intermediate stage after removal of one Li+ and e-.

14
M. Tang, N.N. Bui, J. Zheng et al. Journal of Energy Chemistry 60 (2021) 9–15

This mechanism including TCNQ dimer formation is also supported [15] W. Xiong, W. Huang, M. Zhang, P. Hu, H. Cui, Q. Zhang, Chem. Mater. 31 (2019)
8069–8075.
by our NMR results in a separate study.
[16] T.B. Schon, B.T. McAllister, P.-F. Li, D.S. Seferos, Chem. Soci. Rev. 45 (2016)
6345–6404.
4. Conclusions [17] B. Häupler, R. Burges, T. Janoschka, T. Jähnert, A. Wild, U.S. Schubert, J. Mater.
Chem., A 2 (2014) 8999–9001.
[18] J. Xie, Q. Zhang, J. Mater. Chem. A 4 (2016) 7091–7106.
In situ EPR has been performed on the lithiation and delithiation [19] M. Miroshnikov, K.P. Divya, G. Babu, M. Meiyazhagan, L.M.R. Arava, P.M.
process of organic TCNQ electrodes. The integrated EPR signal Ajayan, G. John, J. Mater. Chem., A 4 (2016) 12370–12386.
[20] C. Han, H. Li, R. Shi, T. Zhang, J. Tong, J. Li, B. Li, J. Mater. Chem., A 7 (2019)
reveals no evidence of radical formation during the early stage of 23378–23415.
lithiation, instead, TCNQ dimers are formed during the first elec- [21] M.C. Grossel, F.A. Evans, J.A. Hriljac, K. Prout, S.C. Weston, J. Chem. Soc., Chem.
tron and lithium insertion. The 2nd electron and lithium insertion Commun. 1 (1990) 1494–1495.
[22] Z.-J. Li, Z.-R. Li, F.-F. Wang, F. Ma, M.-M. Chen, X.-R. Huang, Chem. Phys. Lett.
breaks down the dimers and allows sufficient delocalization of
468 (2009) 319–324.
electrons within the formed benzene anions, which leads to the [23] Y. Jing, Y. Liang, S. Gheytani, Y. Yao, Nano Energy 37 (2017) 46–52.
observed increase in EPR signal. The TCNQ core is perturbed upon [24] B. Grossmann, J. Heinze, T. Moll, C. Palivan, S. Ivan, G. Gescheidt, J. Phys. Chem.
B 108 (2004) 4669–4672.
further cycling, confirmed by the increased maximum EPR signal at
[25] C.E. Tait, P. Neuhaus, M.D. Peeks, H.L. Anderson, C.R. Timmel, J. Am. Chem. Soc.
the end of each lithiation-delithiation cycle. In situ EPR serves as a 137 (2015) 8284–8293.
powerful, sensitive, and convenient tool to follow electron activi- [26] J. Rawson, P.J. Angiolillo, M.J. Therien, P. Natl. Acad. Sci. 112 (2015) 13779–
ties, which in turn affords insights into reaction mechanisms of 13783.
[27] H. Kaftelen, M. Tuncer, S. Tu, S. Repp, H. Goçmez, R. Thomann, S. Weber, E.
organic electrodes in rechargeable batteries in real-time. Erdem, J. Mater. Chem. 1 (2013) 9973–9982.
[28] M.D. Peeks, C.E. Tait, P. Neuhaus, G.M. Fischer, M. Hoffmann, R. Haver, A.
Declaration of Competing Interest Cnossen, J.R. Harmer, C.R. Timmel, H.L. Anderson, J. Am. Chem. Soc. 139 (2017)
10461–10471.
[29] C. Berthier, J.R. Cooper, D. Jerome, G. Soda, C. Weyl, J.M. Fabre, L. Giral, Mol.
The authors declare that they have no known competing finan- Cryst. Liq. Cryst. 32 (1976) 267–270.
cial interests or personal relationships that could have appeared [30] M. Botko, E. Čižmár, M. Kajňaková, A. Feher, Acta Phys. Pol., A 126 (2014) 252–
253.
to influence the work reported in this paper. [31] S. Richert, C.E. Tait, C.R. Timmel, J. Magn. Res. 280 (2017) 103–116.
[32] C. Li, M. Shen, X. Lou, B. Hu, J. Phys. Chem. C 122 (2018) 27224–27232.
Acknowledgments [33] S. Gu, S. Wu, L. Cao, M. Li, N. Qin, J. Zhu, Z. Wang, Y. Li, Z. Li, J. Chen, Z. Lu, J. Am.
Chem. Soc. 141 (2019) 9623–9628.
[34] J. Wandt, C. Marino, H.A. Gasteiger, P. Jakes, R.-A. Eichel, J. Granwehr, Energy
This work was supported by the National Science Foundation Environ. Sci. 8 (2015) 1358–1367.
under Grant No. DMR-1847038. All EPR experiments were carried [35] Q. Wang, J. Zheng, E. Walter, H. Pan, D. Lv, P. Zuo, H. Chen, Z.D. Deng, B.Y. Liaw,
X. Yu, X. Yang, J.-G. Zhang, J. Liu, J. Xiao, J. Electrochem. Soc. 162 (2015) A474–
out at the NHMFL which is supported by the National Science
A478.
Foundation Cooperative Agreement No. DMR-1644779 and the [36] M. Sathiya, J.-B. Leriche, E. Salager, D. Gourier, J.-M. Tarascon, H. Vezin, Nat.
State of Florida. Commun. 6 (2015) 6276.
[37] J.B. Priebe, M. Karnahl, H. Junge, M. Beller, D. Hollmann, A. Brückner, Angew.
Chem. Int. Ed. 5 (2013) 11420–11424.
References [38] J. Rabeah, U. Bentrup, R. Stçßer, A. Brückner, Angew. Chem. Int. Ed. 54 (2015)
11791–11794.
[1] J.-M. Tarascon, M. Armand, Nature 414 (2001) 359–367. [39] J. Wandt, P. Jakes, J. Granwehr, R.-A. Eichel, H.A. Gasteiger, Mater. Today 21
[2] M. Armand, J.-M. Tarascon, Nature 451 (2008) 652–657. (2017) 231–240.
[3] Y. Wen, X. Chen, X. Lu, L. Gu, J. Energy Chem. 27 (2018) 1397–1401. [40] M. Tang, A. Dalzini, X. Li, X. Feng, P.-H. Chien, L. Song, Y.-Y. Hu, J. Phys. Chem.
[4] N. Zhang, X. Zhang, E. Shi, S. Zhao, K. Jiang, D. Wang, P. Wang, S. Guo, P. He, H. Lett. (2017) 4009–4016.
Zhou, J. Energy Chem. 27 (2018) 1655–1660. [41] J. Wang, M. Yang, C. Zhao, B. Hu, X. Lou, F. Geng, W. Tong, B. Hu, C. Li, Phys.
[5] Y. Chen, Y. Kang, Y. Zhao, L. Wang, J. Liu, Y. Li, Z. Liang, X. He, X. Li, N. Tavajohi, Chem. Chem. Phys. 21 (2019) 24017–24025.
B. Li, J. Energy Chem. (2020) doi.org/10.1016/j.jechem.2020.10.017 [42] S. Stoll, A. Schweiger, J. Magn. Reson. 178 (2006) 42–55.
[6] Y. Hanyu, I. Honma, Sci. Rep. 2 (2012) 453–458. [43] A. Ishitani, S. Nagakura, Mol. Phys. 12 (1967) 1–12.
[7] Y. Liang, Y. Jing, S. Gheytani, K.-Y. Lee, P. Liu, A. Facchetti, Y. Yao, Nat. Mater. 16 [44] J. Krzystek, A. Ozarowski, J. Telser, Coord. Chem. Rev. 250 (2006) 2308–2324.
(2017) 841–848. [45] A.L. Michan, M. Leskes, C.P. Grey, Chem. Mater. 28 (2016) 385–398.
[8] Y. Lu, Q. Zhang, L. Li, Z. Niu, J. Chen, Chemistry 4 (2018) 2786–2813. [46] Y. Jin, N.-J.-H. Kneusels, P.C.M.M. Magusin, G. Kim, E. Castillo-Martínez, L.E.
[9] Z. Song, Y. Qian, X. Liu, T. Zhang, Y. Zhu, H. Yu, M. Otani, H. Zhou, Energy Marbella, R.N. Kerber, D.J. Howe, S. Paul, T. Liu, C.P. Grey, J. Am. Chem. Soc. 139
Environ. Sci. 7 (2014) 4077–4086. (2017) 14992–15004.
[10] X. Chi, Y. Liang, F. Hao, Y. Zhang, J. Whiteley, H. Dong, P. Hu, S. Lee, Y. Yao, [47] N.M. Trease, L. Zhou, H.J. Chang, B.Y. Zhu, C.P. Grey, Solid State Nucl. Magn.
Angew. Chem. Int. Ed. 57 (2018) 2630–2634. Reson. 42 (2012) 62–70.
[11] W. Huang, Z. Zhu, L. Wang, S. Wang, H. Li, Z. Tao, J. Shi, L. Guan, J. Chen, Angew. [48] L. Zhou, M. Leskes, A.J. Ilott, N.M. Trease, C.P. Grey, J. Magn. Reson. 234 (2013)
Chem. Int. Ed. 52 (2013) 9162–9166. 44–57.
[12] W. Huang, S. Zheng, X. Zhang, W. Zhou, W. Xiong, J. Chen, Energy Storage [49] F.J. Dyson, Phys. Rev. 98 (1955) 349–359.
Mater. 26 (2020) 465–471. [50] G. Feher, A.F. Kip, Phys. Rev. 98 (1955) 337–348.
[13] K. Taniguchi, K. Narushima, J. Mahin, W. Kosaka, H. Miyasaka, Angew. Chem. [51] A.L. Sutton, B.F. Abrahams, D.M. D’Alessandro, T.A. Hudson, R. Robson, P.M.
Int. Ed. 55 (2016) 5238–5242. Usov, Cryst. Eng. Comm. 18 (2016) 8906–8914.
[14] C. Fang, Y. Huang, L. Yuan, Y. Liu, W. Chen, Y. Huang, K. Chen, J. Han, Q. Liu, Y. [52] Y. Chen, S. Manzhos, Phys. Chem. Chem. Phys. 18 (2016) 1470–1477.
Huang, Angew. Chem. Int. Ed. 56 (2017) 6793–6797.

15