www.sciencemag.org/content/361/6404/777/suppl/DC1

Supplementary Material for

A high-energy-density lithium-oxygen battery based on a reversible four-electron conversion to lithium oxide

C. Xia, C. Y. Kwok, L. F. Nazar*

*Corresponding author. Email: lfnazar@uwaterloo.ca

Published 24 August 2018, Science 361, 777 (2017)

DOI: 10.1126/science.aas9343

This PDF file includes:

Materials and Methods Figs. S1 to S11 Table S1 References

2 | P a g e

Materials and Methods

Preparation of LiNO3-KNO3 molten salt electrolyte

The molten nitrate salt electrolyte was composed of lithium nitrate (99.99 %, Sigma-

Aldrich) and potassium nitrate (≥99.0%, ReagentPlus®, Sigma-Aldrich) with a mole ratio

of 42:58, which forms a eutectic. Since the glass fiber separator (Whatman) was poorly

wet by the molten nitrate, the electrolyte filled separator was prepared by immersing the

glass fiber paper in an aqueous solution of LiNO3-KNO3 at a salt concentration of 0.25

g∙mL-1. The wetted glass fiber discs with a diameter of 12 mm were placed in a 180 ºC

oven for 20 min to evaporate water and allow the absorbed nitrates to melt to form a molten

salt. This process was iterated three times until all of the pores in the glass fiber were filled.

The mass loading of the electrolyte was determined to be 100 ± 10 mg per glass fiber. After

the electrolyte-filled separators were vacuum-dried at 200 oC for one day in a Büchi glass

oven, they were transferred into a glovebox for cell assembly.

Preparation of Li1.5Al0.5Ge1.5(PO4)3 (LAGP) solid electrolyte

LAGP was synthesized via a solid-state reaction. First, 2.65 g Li2CO3 (≥99.0%, ACS

reagent), 1.22 g Al2O3 (Sigma-Aldrich), 7.51 g GeO2 (≥99.99%, Sigma-Aldrich), and 16.52

g NH4H2PO4 (99.999%, Sigma-Aldrich) precursors were ball-milled with acetone at 250

rpm for 4 h using agate mortar balls (6 mm, 5 g). The obtained mixture was then dried in

a 60 °C oven for 12 h to remove the acetone. The white powders were then calcined at 600

ºC in air for 4 h to decompose the precursors. The subsequent light grey powders were ball-

milled at 250 rpm for 4 h, followed by pelletization at an applied pressure of 3 tons. The

square pellet (1 cm2, 4 mm) was then calcined at 900 ºC for 6 h. Another ball-milling

procedure was employed to grind the pellets, and the resulting particles were sieved to a

dimension smaller than 106 μm in diameter. The solid electrolyte powder (1 g) was pressed

with 5 tons of pressure for 1 min. The resulting white disc was sintered at 900 ºC for 10 h,

with programmed heating and cooling rates of 1 ºC·min-1.

Preparation of the Ni-nitrate composite cathode

Typically, 4 g Ni powder (Anachemia) was mixed with 4 mL of LiNO3-KNO3 aqueous

solution (0.25 g∙mL-1). The mixture was then heated in an oven at 180 ºC for 20 min to

remove water and form a thin layer of the molten nitrate on the Ni nanoparticles and 0.2 g

of the composite cathode powder were pressed onto a stainless steel current collector

(SS316) under an applied pressure of 1 ton. The geometric area of the cathode was 1 cm2.

The cathodes were further dried at 200 oC under vacuum in a Büchi glass oven for one day

prior to use. Characterization of the composite cathode is shown in Fig. S3.

Preparation of Li2O2-prefilled Ni-nitrate electrode

In an Ar-filled glovebox, the prefilled electrode was prepared by grinding and mixing 200

mg of the Ni-nitrate composite material and 50 mg of the commercial Li2O2 powder

(technical grade, 90.0%, Sigma-Aldrich) via a mortar. The mixture powder was then

pressed onto a stainless steel current collector (SS316) under an applied pressure of 1 ton.

The geometric area of the electrode was 1 cm2.

3 | P a g e

GC-MS analysis for the chemical stability of Li2O and Li2O2 in DMSO

Dimethyl sulfoxide (DMSO, anhydrous, ≥99.9 %, Sigma-Aldrich) was distilled under

partial pressure and stored over molecular sieves (3 Å, beads 8 – 12 mesh, Sigma-Aldrich)

for two days prior to use. In an Ar-filled glovebox, commercial Li2O and Li2O2 powders

were mixed with 6.2 mL of the purified DMSO. The stoichiometric mole ratio between the

commercial powders and DMSO was kept at 1:100. The resulting suspensions were stirred

using a magnetic stir bar for 300 hours, followed by centrifugation (10 kRPM, 10min) to

remove the solids. The supernatant was collected and further diluted in acetonitrile (1:10

vol. %) before subject to GC-MS analysis. Single-ion monitoring mode was utilized to

enhance the sensitivity of dimethyl sulfone signal.

Preparation of SuperP carbon cathode

The cathode slurry consisted of 85 mg SuperP carbon (Toyota) and 15 mg

polytetrafluorgoethylene (PTFE, Sigma-Aldrich) was suspended in 10 mL ethanol by

magnetic stirring for 4 h. The mixture was sprayed onto a gas diffusion layer (GDL H2315,

Freudenberg) as a support. Cathodes (geometric area of 1 cm2) were punched from the

GDL with a typical carbon loading of 2 ± 0.5 mg. After drying at ambient conditions for 1

h, 0.25 g∙mL-1 LiNO3-KNO3 aqueous solution was sprayed on the cathodes which were

then dried at 200 °C under vacuum in a Büchi glass oven for one day prior to use. The

loading of nitrate on the carbon cathode was 25 ± 5 mg.

Preparation of organic electrolyte

Tetraethylene glycol dimethyl ether (TEGDME, 99%, Sigma-Aldrich) was distilled over

calcium hydride (reagent grade, 95%, Sigma-Aldrich) under a N2 atmosphere and stored

over molecular sieves (4 Å, beads 8 – 12 mesh, Sigma-Aldrich) for one week prior to use.

Lithium bis(trifluoromethane) sulfonamide (LiTFSI, 99.95 %, Sigma-Aldrich) was

vacuum dried at 150 °C overnight in a Büchi glass oven. An electrolyte solution containing

0.5 M LiTFSI in TEGDME was prepared in an argon-filled glovebox with a water and

oxygen content less than 0.5 and 1.0 ppm, respectively. The water content of the electrolyte

was confirmed to be < 10 ppm, as determined by a Karl Fischer titrator (Mettler Toledo).

Cell assembly and galvanostatic cycling

SwagelokTM-type stainless steel cells were used for battery testing which were assembled

in an argon-filled glovebox. As illustrated in Fig. 1B, the cell was assembled by stacking

a lithium anode, the electrolyte-filled glass fiber, the LAGP solid electrolyte, and the

composite cathode. After purging the cell with pure O2 (5.0 Research, Prixair) for 10 s, it

was sealed under 3 atm oxygen. The cell was placed in an oven at 150 °C for 2 h under

open circuit conditions. The electrochemical performance of cells was evaluated using a

multi-channel Arbin instrument at a current density of 0.1 mA·cm-2 with a potential

window between 2.6 and 3.5 V vs. Li/Li+.

Physical characterization

Powder XRD was performed on a Bruker D8-Advance powder diffractometer equipped

with a Vantec-1 detector (Cu Kα radiation). The XRD patterns of the samples were

collected using a gastight sample holder with a Kapton film window. UV-vis spectroscopy

was performed on a Cary 300 Bio UV-visible spectrometer. Raman spectroscopy was

4 | P a g e

carried out on a Raman HORIBA HR800 equipped with a green laser (λ = 514 nm). To

measure gas evolution during cell operation, the Swagelok®-type cells were monitored with

an on-line mass spectrometer (OEMS) system (RGA 200 Stanford Research Systems). A

continuous flow of Ar (5.0, Praxair) was used to sweep the gases evolved during the

electrochemical process across a 50 mm diameter capillary. XPS analysis was performed

on a Thermo VG Scientific ESCALAB 250. The samples were transferred to an XPS

microprobe without air exposure. All spectra were fitted with Gaussian-Lorentzian

functions and a Shirley-type background using CasaXPS software, and were calibrated

with the C 1s photoemission peak at 285.0 eV. The Ni 2p doublet peaks were fit using

equal full width at half maximum (FWHM), with a 2:1 area ratio and splitting of 17.5 eV

between the Ni 2p3/2 and Ni 2p1/2, spin doublets respectively. Multiplet components were

not considered when fitting the Ni 2p spectrum. SEM studies were carried out on a Zeiss

Ultra field emission SEM instrument, and TEM was performed on a ZEISS Libra 200 MC

instrument operating at 200 kV. GC-MS analysis was performed on an Agilent 5975B

GC/MS.

5 | P a g e

S1. Chemical stability of Li2O and Li2O2 in DMSO

Fig. S1. Quantitative GC-MS results of the reaction of DMSO with Li2O (red), and Li2O2 (blue) for 300 h,

compared to neat DMSO (black). The GC trace shows the eluted fraction from the column that corresponds

to DMSO2, as proven by MS analysis (inset).

DMSO was chosen as a representative organic solvent to compare the chemical reactivity

of Li2O with that of Li2O2. DMSO was contacted with the same quantity of Li2O or Li2O2

for 300 hours, and a standard aliquot was extracted and subjected to GC-MS analysis. The

results show that the same fraction of oxidized DMSO (DMSO2; (m/z = 94) (33) is present

after reaction with Li2O as that in neat DMSO where it is present in trace amounts,

indicating that Li2O does not oxidize the solvent. In contrast, a significant quantity of

DMSO2 is observed on reaction of DMSO with Li2O2, in complete accord with the results

of Shao-horn et al.(9) on the reactivity of Li2O2 with DMSO.

6 | P a g e

S2. Comparison of product crossover with and without the LAGP solid electrolyte

Fig. S2. Comparison of product crossover with and without the LAGP solid electrolyte. (A) Schematic

of a molten salt Li-O2 cell without LAGP electrolyte. (B) Photograph of cuvettes filled with solutions

subjected to a Griess analysis to test for LiNO2; the solution were extracted from the different layers of the

cell with two glass fiber separators (a,b) in place. (C) The corresponding UV-vis spectra of the cuvette

solutions: the Griess test results in formation of a purple azo-dye in the presence of NO2-, which exhibits a

strong absorbance at 541 nm (see section S6.4). The cell was maintained at open circuit conditions (OCV)

for 2 h prior to disassembly for analysis. (D) Comparison of cathodic products in the (black) absence and

(red) presence of the LAGP solid electrolyte at OCV after 64 h. Black→ both separators (a,b) were glass

fiber membranes; Red→one membrane (b) was LAGP. The operating temperature was 150 °C.

Fig. S2A shows the configuration of a molten salt Li-O2 cell with two layers of electrolyte-

filled glass fiber separators (a,b). After placing the cell at 150 oC under OCV conditions

for 2 h, it was disassembled in a glovebox. Due to the reaction between the lithium anode

7 | P a g e

and molten nitrate (Eq. S1), Li2O and LiNO2 were formed on the surface of lithium and

subsequently solubilized in the molten nitrate electrolyte. The LiNO2 content in each layer

was quantified using UV-vis spectroscopy (for details see section S6-4). The diffusion of

these species led to an uneven distribution of LiNO2 in each of the cell components as

visually marked by the Griess test for LiNO2 in Fig. S2B and C. In addition, Li2O and

LiNO2 reacted with the carbon cathode forming a Li2CO3 side product (Eq. S2 and S3).

Analysis of Li2O and Li2CO3 is described below in section S6.

To restrict the crossover of these byproducts and the formation of carbonate, a non-

porous LAGP solid electrolyte was used to substitute one separator layer (layer b in Fig.

S2A). The cells were subject to the same conditions as above but held at 150 oC for 64 h.

Fig. S2D shows that the cathode without LAGP protection contains considerable

byproducts, but with the solid LAGP electrolyte in place, the byproducts total < 0.1 μmol.

2Li + LiNO3 → Li2O + LiNO2 ΔrG150° = -467.90 kJ∙mol-1 Eq. S1

Li2O + C + O2 → Li2CO3 ΔrG150° = -551.26 kJ∙mol-1 Eq. S2

Li2O + C + 2LiNO3 → Li2CO3 + 2LiNO2 ΔrG150° = -395.93 kJ∙mol-1 Eq. S3

8 | P a g e

S3. Characterization of the Ni-nitrate composite cathode

Fig. S3. Characterization of the Ni-nitrate composite. (A) SEM image and the corresponding TEM image

(inset). (B) XRD pattern. XPS spectra: (C) Ni 2p, (D) O 1s, and (E) Li 1s. Multiplet components were not

considered when fitting the Ni 2p spectrum and only the lower binding components of the Ni 2p spectrum

are displayed.

As shown in Fig. S3A, the Ni-nitrate composite cathode is composed of Ni nanoparticles

covered with a very thin layer of nitrate melt. XPS analysis revealed that Ni particles in the

composite cathode are covered with two types of oxide species due to the oxidation of Ni

by molten nitrate. The major peak at 855.4 eV in the Ni 2p spectrum (Fig. S3c) is assigned

to Ni2O3 (34), evidenced by its corresponding O 1s peak at 531.9 eV in Fig. S3. In addition,

the formation of lithiated nickel (III) oxide (LixNiO2) contributes to its characteristic Ni 2p

peak at 854.0 eV; O 1s peak at 529.6 eV; and Li 1s peak at 54.9 eV as described in the

literature (35,36). The peak at 56.0 eV in the Li 1s spectrum is residual LiNO3 (37),

attributed to incomplete washing of the sample.

9 | P a g e

S4. XPS analysis of the discharged carbon cathode

Fig. S4. XPS spectra of a discharged carbon cathode: (A) O 1s and (B) Li 1s.

Fig. S4 presents the XPS spectra of a fully discharged (2.6 V) carbon cathode. The peaks

at 531.2 eV in the O 1s spectrum and at 54.5 eV in the Li 1s spectrum are assigned to Li2O2,

consistent with its presence in the XRD pattern (Fig. 2B). Two peaks are also observed at

528.6 eV and 53.6 eV that are characteristic of Li2O (38). As Li2O is not observed in the

XRD pattern, we conclude it is amorphous.

10 | P a g e

S5. Determination of the solubility of Li2O in molten nitrate

Fig. S5. Schematic diagram of the apparatus for measuring the solubility of Li2O in molten nitrate

at 150 °C.

The solubility of Li2O in the molten nitrate was measured using a home-made apparatus

illustrated in Fig. S5. Commercial Li2O powder (1.0 g) was added to the LiNO3-KNO3

eutectic mixture (20 g) and placed in a glass beaker. The eutectic nitrate mixture (4 g) alone

was also placed into a separate PTFE container whose bottom was sealed with a glass fiber

membrane. The apparatus was assembled in an Ar-filled glovebox by inserting the PTFE

container into the glass beaker. The apparatus was heated at 150 ºC under vacuum to melt

the eutectic. A glass pipet was inserted into the PTFE container to extract the molten nitrate

for the solubility measurement, and the amount of Li2O in the eutectic was quantified by

titration (see section S6-3 for details). As shown in Table S1, the solubility of Li2O in

molten nitrate was 27.6 mM at 150 °C.

Table S1. Solubility of lithium oxide in Li/KNO3 molten salt electrolyte at 150 °C

Time 24 h 48 h 72 h Average

Solubility [mM] 30.8 23.3 28.8 27.6 ± 4.0

11 | P a g e

S6. Quantification of the discharge products

Preparation of the discharged cathode sample. After disassembling the cell, the cathode

was removed from the glovebox and transferred into a glass vial filled with 3 mL water.

After 5 min of ultrasonsication, the solution was filtered to remove any insoluble Ni

nanoparticles. The resulting transparent solution was used for the quantification of Li2CO3,

Li2O2, Li2O, and LiNO2.

1. Li2CO3

Theory. Lithium carbonate reacts with HCl to form CO2 (Eq. S4), which was detected by

a mass-spectrometer (MS).

Li2CO3 + 2HCl = 2LiCl + H2O + CO2 Eq. S4

Calibration. The amount of CO2 was calibrated by diluting an Ar/CO2 gas mixture of

known concentration (2000 ppm CO2, Praxair) with an additional Ar carrier gas prior to

MS detection. As shown in Fig. S6-1A, the amount of CO2 was correlated to the intensity

of the MS signal. The total gas flow was maintained at 5 mL∙min-1. By integrating the MS

signal area, the total number of moles of CO2 (n𝐿𝑖2𝐶𝑂3) was calculated using the standard

curve (Fig. S6-1B) and Eq. S4.

Fig. S6-1. Quantification of CO2. (A) Response of the MS signals at different CO2 concentrations and (B)

the corresponding standard curve.

Experimental. The cathodic solution (0.5 mL) was first introduced into a titration cell

under Ar flow, followed by an injection of HCl (0.1M, 1 mL) into the cell using a syringe.

The evolved CO2 which was detected by the mass-spectrometer.

2. Li2O2

Theory: The quantification of lithium peroxide was accomplished by first hydrolyzing

Li2O2 to form LiOH and H2O2 (Eq. S5), followed by oxidation of TiOSO4 by hydrogen

peroxide (Eq. S6). The final yellow colored product (TiO2SO4) has an adsorption peak at

405 nm (39), which was detected by an UV-vis spectrometer.

Li2O2 + 2H2O → 2LiOH + H2O2 Eq. S5

H2O2 + TiOSO4 → TiO2SO4 + H2O Eq. S6

12 | P a g e

Calibration. To establish a standard curve, different concentrations of H2O2 were prepared

by diluting a stock solution of H2O2 (30 wt. % in H2O, Sigma-Aldrich) in water. The precise

concentration of H2O2 was determined by KMnO4 titration in the presence of H2SO4. The

diluted H2O2 (0.5 mL) and TiOSO4 (0.5 mL, 15 wt. % in dilute sulfuric acid, Sigma-

Aldrich) were added to a 1.5 mL PMMA cuvette and the UV-vis spectrum was recorded.

As shown in Fig. S6-2A, a peak at 405 nm corresponding to TiO2SO4 was detected at

different concentrations of H2O2, was used to establish a standard curve (Fig. S6-2B).

Fig. S6-2. Quantification of H2O2. (A) UV-vis spectrum of TiO2SO4 complex prepared with different

concentrations of H2O2 and (B) the corresponding standard curve.

Quantification. Similar to the standard solution, the cathodic solution (0.5 mL) and

TiOSO4 solution (0.5 mL) were introduced in a PMMA cuvette for UV-vis spectroscopic

analysis.

3. Li2O

Theory. The quantification of lithium oxide was accomplished by first hydrolyzing Li2O

to form LiOH (Eq. S7) that was titrated with HCl (Eq. S8). Contribution to the total

alkalinity (n𝑎𝑙𝑙) of the solution was via hydrolysis of three components: Li2CO3 (n𝐿𝑖2𝐶𝑂3),

Li2O2 (n𝐿𝑖2𝑂2), and Li2O (n𝐿𝑖2𝑂). Prior knowledge of the total amount of Li2CO3 and Li2O2

in the solution enables us to calculate the total amount of Li2O as shown in Eq. S9

Li2O + H2O = 2LiOH Eq. S7

LiOH + HCl = LiCl2 + H2O Eq. S8

n𝐿𝑖2𝑂 =1

2(n𝑎𝑙𝑙 − 2n𝐿𝑖2𝑂2

− 2n𝐿𝑖2𝐶𝑂3) Eq. S9

Calibration. A series of HCl standard solutions were prepared by diluting a stock HCl

solution (37 wt. %, Sigma-Aldrich) in deionized water. The concentration of HCl in the

standard solutions was determined by titration using Na2CO3 (≥ 99.5 % anhydrous, Sigma-

Aldich). The endpoint of titration was determined by the color change of a double indicator

(Bromocresol Green/Methyl Red, Sigma-Aldrich) from blue to pink.

Experimental. The total alkalinity (n𝑎𝑙𝑙) in the cathode solution (1 mL) was obtained by

titrating it with the HCl standard solution until the indicator changed color.

13 | P a g e

4. LiNO2

Theory. The nitrite concentration was quantified by a Griess test illustrated in Fig. S6-4A.

Nitrite reacts with sulfanilic acid to form a diazonium salt (step 1), which was then reacts

with a Griess reagent (N-(1-naphthyl)ethylenediamine dihydrochloride) to form a purple-

colored azo dye (step 2) which has an absorbance at 541 nm (40).

Fig. S6-4. Quantification of LiNO2. (A) Schematic diagram for the nitrite quantification via Griess reaction.

(B) UV-vis spectra of the purple azo dye solution prepared at different concentrations of NaNO2 and (C) the

corresponding standard curve.

Calibration. To establish a standard curve, different concentrations of aqueous NaNO2

(0.25 mL, 99.999 %, Sigma-Aldrich) solution and sulfanilic acid (0.5 mL, ≥99.0 %, Sigma-

Aldrich) were mixed in a 1.5 mL PMMA cuvette. A N-(1-naphthyl)ethylenediamine

dihydrochloride aqueous solution (0.25 mL) was then introduced to the same cuvette to

afford a purple colored solution. As shown in Fig. S6-4B, a peak at 541 nm corresponding

to the azo dye was detected at different concentrations of NaNO2, (Fig. S6-4C) to establish

a standard curve via the Beer-Lambert law.

Quantification of LiNO2. The cathodic solution (0.25 ml) used for quantifying LiNO2

followed the same procedure described in the calibration step.

14 | P a g e

S7. Electrochemical reversibility of the carbon cathode

Fig. S7 Charge profile and the corresponding MS results of the Li-O2 cells using carbon cathodes pre-filled

with (A) Li2O and (B) Li2CO3. The charge current was 0.1 mA·cm-2.

To examine the electrochemical reversibility of Li2O and Li2CO3 on a carbon cathode,

SuperP carbon cathodes were pre-filled with either 1 mg of commercial Li2O or Li2CO3

powders. Gaseous products were monitored by an online MS while charging the cells using

these pre-filled cathodes. As shown in Fig. S7A, the charge profile of the cell pre-filled

with Li2O presented two voltage plateaus at 3 and 3.75 V. Although oxygen did not evolve,

signals corresponding to CO2 and NO were identified at potentials above 3.4 V. In contrast

to the chemically stable Ni-nitrate composite cathode, the carbon cathode reacted with

lithium oxide and the molten-nitrate to form Li2CO3 and LiNO2 (Eq. 3 and 4). The cell pre-

filled with Li2CO3 exhibited a charge plateau at 4 V (Fig. S7B). Similarly, oxygen was not

detected but CO2 and NO signals were observed. This suggests that the molten nitrate

electrolyte decomposed in parallel to the oxidation of Li2CO3. Hence, the formation of both

Li2O and Li2CO3 is electrochemically irreversible on a carbon cathode.

15 | P a g e

S8. Passivation of the carbon cathode by lithium carbonate

Fig. S8. XRD pattern (A) and SEM image (B) of a carbon cathode after cell degradation. Crystalline needles

of Li2CO3 were formed, which passivated the carbon electrode.

16 | P a g e

S9. Cycling performance of Li-O2 cell with Ni-nitrate cathode at a current density

of 0.1 mA·cm-2

Fig. S9. Cycling performance of a molten salt electrolyte Li-O2 cell with the Ni-nitrate composite cathode at

an applied current of 0.1 mA∙cm-2, showing a similar profile as full discharge (Fig 2A) recorded at the same

current density. The polarization (0.16 V) after the first cycle is slightly lower than in Fig 2A (0.2 V) since

here, the capacity is limited.

17 | P a g e

S10. Degradation of LAGP on cycling

Fig. S10. Visual inspection of the anodic side of the LAGP electrolyte layer after cell cycling. (A)

Photograph showing black spots and cracks on the surface of the membrane and (B) SEM image of a black

spot showing its porous structure. (C) Impedance spectra of the LAGP (black) before and (red) after cycling

for 50 cycles.

C

0 100 200 300 4000

100

200

300

-Z''

(.c

m2)

Z' (.cm2)

18 | P a g e

S11. Analysis of Li2O2 in the charge cathode pre-filled with Li2O

Fig. S11 (A) Charge profile of a Li-O2 cell with a Ni-nitrate composite cathode pre-filled with commercial

Li2O. The charge cut-off was set at a capacity of 0.4 mA·h·cm-2 (B) UV-vis spectra of the corresponding

charged cathode indicates the absence of Li2O2. The charge current density was 0.2 mA·cm-2

To examine the possibility of formation of Li2O2 as an intermediate upon charge, a molten

electrolyte cell using a Ni-nitrate composite cathode pre-filled with commercial Li2O

powder was charged to 0.4 mA·h·cm-2 (Fig. S11 A). After charge, the cell was rapidly

disassembled in a glovebox and the peroxide content in the charged cathode was analyzed

by UV-vis spectroscopy as demonstrated in S6. The lack of a peak at 408 nm indicates that

Li2O2 is not formed on charging, but rather Li2O is directly oxidized to O2 (Fig. S11B).

References and Notes 1. Z. Peng, S. A. Freunberger, Y. Chen, P. G. Bruce, A reversible and higher-rate Li-O2 battery.

Science 337, 563–566 (2012). doi:10.1126/science.1223985 Medline 2. Y. C. Lu, B. M. Gallant, D. G. Kwabi, J. R. Harding, R. R. Mitchell, M. S. Whittingham, Y.

Shao-Horn, Lithium-oxygen batteries: Bridging mechanistic understanding and battery performance. Energy Environ. Sci. 6, 750–768 (2013). doi:10.1039/c3ee23966g

3. P. Hartmann, C. L. Bender, M. Vračar, A. K. Dürr, A. Garsuch, J. Janek, P. Adelhelm, A rechargeable room-temperature sodium superoxide (NaO2) battery. Nat. Mater. 12, 228–232 (2013). doi:10.1038/nmat3486 Medline

4. D. Sharon, D. Hirshberg, M. Afri, A. Garsuch, A. A. Frimer, D. Aurbach, Lithium-oxygen electrochemistry in non-aqueous solutions. Isr. J. Chem. 55, 508–520 (2015). doi:10.1002/ijch.201400135

5. D. Aurbach, B. D. McCloskey, L. F. Nazar, P. G. Bruce, Advances in understanding mechanisms underpinning lithium-air batteries. Nat. Energy 1, 16128 (2016). doi:10.1038/nenergy.2016.128

6. J. Lu, Y. J. Lee, X. Luo, K. C. Lau, M. Asadi, H.-H. Wang, S. Brombosz, J. Wen, D. Zhai, Z. Chen, D. J. Miller, Y. S. Jeong, J.-B. Park, Z. Z. Fang, B. Kumar, A. Salehi-Khojin, Y.-K. Sun, L. A. Curtiss, K. Amine, A lithium-oxygen battery based on lithium superoxide. Nature 529, 377–382 (2016). doi:10.1038/nature16484 Medline

7. T. Liu, M. Leskes, W. Yu, A. J. Moore, L. Zhou, P. M. Bayley, G. Kim, C. P. Grey, Cycling Li-O2 batteries via LiOH formation and decomposition. Science 350, 530–533 (2015). doi:10.1126/science.aac7730 Medline

8. C. M. Burke, R. Black, I. R. Kochetkov, V. Giordani, D. Addison, L. F. Nazar, B. D. McCloskey, Implication of 4 e– reduction via iodine redox mediation in Li-O2 batteries. ACS Energy Lett. 1, 747–756 (2016). doi:10.1021/acsenergylett.6b00328

9. D. G. Kwabi, T. P. Batcho, C. V. Amanchukwu, N. Ortiz-Vitoriano, P. Hammond, C. V. Thompson, Y. Shao-Horn, Chemical instability of dimethyl sulfoxide in lithium-air batteries. J. Phys. Chem. Lett. 5, 2850–2856 (2014). doi:10.1021/jz5013824 Medline

10. R. Black, S. H. Oh, J. H. Lee, T. Yim, B. Adams, L. F. Nazar, Screening for superoxide reactivity in Li-O2 batteries: Effect on Li2O2/LiOH crystallization. J. Am. Chem. Soc. 134, 2902–2905 (2012). doi:10.1021/ja2111543 Medline

11. C. L. Bender, P. Hartmann, M. Vračar, P. Adelhelm, J. Janek, On the thermodynamics, the role of the carbon cathode, and the cycle life of the sodium superoxide (NaO2) battery. Adv. Energy Mater. 4, 1301863 (2014). doi:10.1002/aenm.201301863

12. V. Giordani, D. Tozier, H. Tan, C. M. Burke, B. M. Gallant, J. Uddin, J. R. Greer, B. D. McCloskey, G. V. Chase, D. Addison, A molten salt lithium–oxygen battery. J. Am. Chem. Soc. 138, 2656–2663 (2016). doi:10.1021/jacs.5b11744 Medline

13. N. B. Aetukuri, B. D. McCloskey, J. M. García, L. E. Krupp, V. Viswanathan, A. C. Luntz, Solvating additives drive solution-mediated electrochemistry and enhanced toroid growth

in non-aqueous Li-O2 batteries. Nat. Chem. 7, 50–56 (2015). doi:10.1038/nchem.2132 Medline

14. C. Xia, R. Black, R. Fernandes, B. Adams, L. F. Nazar, The critical role of phase-transfer catalysis in aprotic sodium oxygen batteries. Nat. Chem. 7, 496–501 (2015). doi:10.1038/nchem.2260 Medline

15. V. A. Garten, D. E. Weiss, The ion- and electron-exchange properties of activated carbon in relation to its behavior as a catalyst and adsorbent. Rev. Pure Appl. Chem. 7, 69–122 (1957).

16. B. D. McCloskey, A. Speidel, R. Scheffler, D. C. Miller, V. Viswanathan, J. S. Hummelshøj, J. K. Nørskov, A. C. Luntz, Twin problems of interfacial carbonate formation in nonaqueous Li-O2 batteries. J. Phys. Chem. Lett. 3, 997–1001 (2012). doi:10.1021/jz300243r Medline

17. P. Hartmann, M. Heinemann, C. L. Bender, K. Graf, R.-P. Baumann, P. Adelhelm, C. Heiliger, J. Janek, Discharge and charge reaction paths in sodium–oxygen batteries: Does NaO2 form by direct electrochemical growth or by precipitation from solution? J. Phys. Chem. C 119, 22778–22786 (2015). doi:10.1021/acs.jpcc.5b06007

18. B. D. McCloskey, D. S. Bethune, R. M. Shelby, T. Mori, R. Scheffler, A. Speidel, M. Sherwood, A. C. Luntz, Limitations in rechargeability of Li-O2 batteries and possible origins. J. Phys. Chem. Lett. 3, 3043–3047 (2012). doi:10.1021/jz301359t Medline

19. M. J. Welland, K. C. Lau, P. C. Redfern, L. Liang, D. Zhai, D. Wolf, L. A. Curtiss, An atomistically informed mesoscale model for growth and coarsening during discharge in lithium-oxygen batteries. J. Chem. Phys. 143, 224113 (2015). doi:10.1063/1.4936410 Medline

20. B. D. Adams, R. Black, Z. Williams, R. Fernandes, M. Cuisinier, E. J. Berg, P. Novak, G. K. Murphy, L. F. Nazar, Towards a stable organic electrolyte for the lithium oxygen battery. Adv. Energy Mater. 5, 1400867 (2015). doi:10.1002/aenm.201400867

21. C. Xia, R. Fernandes, F. H. Cho, N. Sudhakar, B. Buonacorsi, S. Walker, M. Xu, J. Baugh, L. F. Nazar, Direct evidence of solution-mediated superoxide transport and organic radical formation in sodium-oxygen batteries. J. Am. Chem. Soc. 138, 11219–11226 (2016). doi:10.1021/jacs.6b05382 Medline

22. W. Chia-Ching, Y. Cheng-Fu, Investigation of the properties of nanostructured Li-doped NiO films using the modified spray pyrolysis method. Nanoscale Res. Lett. 8, 33 (2013). doi:10.1186/1556-276X-8-33 Medline

23. P. Hartmann, T. Leichtweiss, M. R. Busche, M. Schneider, M. Reich, J. Sann, P. Adelhelm, J. Janek, Degradation of NASICON-Type materials in contact with lithium metal: Formation of mixed conducting interphases (MCI) on solid electrolytes. J. Phys. Chem. C 117, 21064–21074 (2013). doi:10.1021/jp4051275

24. B. D. Adams, C. Radtke, R. Black, M. L. Trudeau, K. Zaghib, L. F. Nazar, Current density dependence of peroxide formation in the Li-O2 battery and its effect on charge. Energy Environ. Sci. 6, 1772–1778 (2013). doi:10.1039/c3ee40697k

25. N. Markovic, Surface science studies of model fuel cell electrocatalysts. Surf. Sci. Rep. 45, 117–229 (2002). doi:10.1016/S0167-5729(01)00022-X

26. C. M. Sánchez-Sánchez, A. J. Bard, Hydrogen peroxide production in the oxygen reduction reaction at different electrocatalysts as quantified by scanning electrochemical microscopy. Anal. Chem. 81, 8094–8100 (2009). doi:10.1021/ac901291v Medline

27. J. Suntivich, H. A. Gasteiger, N. Yabuuchi, H. Nakanishi, J. B. Goodenough, Y. Shao-Horn, Design principles for oxygen-reduction activity on perovskite oxide catalysts for fuel cells and metal-air batteries. Nat. Chem. 3, 546–550 (2011). doi:10.1038/nchem.1069 Medline

28. Y. Koyama, T. Mizoguchi, H. Ikeno, I. Tanaka, Electronic structure of lithium nickel oxides by electron energy loss spectroscopy. J. Phys. Chem. B 109, 10749–10755 (2005). doi:10.1021/jp050486b Medline

29. S. Srinivasan, Fuel Cells: From Fundamentals to Applications (Springer, 2010). 30. J. Suntivich, K. J. May, H. A. Gasteiger, J. B. Goodenough, Y. Shao-Horn, A perovskite

oxide optimized for oxygen evolution catalysis from molecular orbital principles. Science 334, 1383–1385 (2011). doi:10.1126/science.1212858 Medline

31. Z. W. Seh, J. Kibsgaard, C. F. Dickens, I. Chorkendorff, J. K. Nørskov, T. F. Jaramillo, Combining theory and experiment in electrocatalysis: Insights into materials design. Science 355, eaad4998 (2017). doi:10.1126/science.aad4998 Medline

32. M. E. Scofield, H. Liu, S. S. Wong, A concise guide to sustainable PEMFCs: Recent advances in improving both oxygen reduction catalysts and proton exchange membranes. Chem. Soc. Rev. 44, 5836–5860 (2015). doi:10.1039/C5CS00302D Medline

33. S. E. Stein, Mass Spectra in NIST Chemistry WebBook, NIST Standard Reference Database Number 69 (National Institute of Standards and Technology Mass Spec Data Center, 2017); https://webbook.nist.gov/chemistry.

34. K. S. Kim, N. Winograd, X-ray photoelectron spectroscopic studies of nickel oxide surface using oxygen and argon ion-bombardment. Surf. Sci. 43, 625–643 (1974). doi:10.1016/0039-6028(74)90281-7

35. A. W. Moses, H. G. G. Flores, J.-G. Kim, M. Langell, Surface properties of LiCoO2, LiNiO2 and LiNi1-xCoxO2. Appl. Surf. Sci. 253, 4782–4791 (2007). doi:10.1016/j.apsusc.2006.10.044

36. A. N. Mansour, Characterization of LiNiO2 by XPS. Surf. Sci. Spectra 3, 279–286 (1994). doi:10.1116/1.1247757

37. S. Aduru, S. Contarini, J. W. Rabalais, Electron-, x-ray-, and ion-stimulated decomposition of nitrate salts. J. Phys. Chem. 90, 1683–1688 (1986). doi:10.1021/j100399a045

38. K. P. C. Yao, D. G. Kwabi, R. A. Quinlan, A. N. Mansour, A. Grimaud, Y.-L. Lee, Y.-C. Lu, Y. Shao-Horn, Thermal stability of Li2O2 and Li2O for Li-air batteries: In situ XRD and XPS studies. J. Electrochem. Soc. 160, A824–A831 (2013). doi:10.1149/2.069306jes

39. X. Gao, Y. Chen, L. Johnson, P. G. Bruce, Promoting solution phase discharge in Li-O2 batteries containing weakly solvating electrolyte solutions. Nat. Mater. 15, 882–888 (2016). doi:10.1038/nmat4629 Medline

40. D. Giustarini, R. Rossi, A. Milzani, I. Dalle-Donne, Nitrite and nitrate measurement by Griess reagent in human plasma: Evaluation of interferences and standardization. Methods Enzymol. 440, 361–380 (2008). doi:10.1016/S0076-6879(07)00823-3 Medline