Direct-contact micro-electrical measurement of the electrical resistivity

2 days ago - ... of the SEI layer is elucidated by calculating the band structures of various LixF compounds and interpreting their effects on the res...
0 downloads 0 Views 3MB Size
Subscriber access provided by UNIVERSITY OF TECHNOLOGY SYDNEY

Communication

Direct-contact micro-electrical measurement of the electrical resistivity of a solid electrolyte interface Jun-Hyoung Park, Yong-Seok Choi, Hyun-Jeong Lee, Hyungcheoul Shim, Jae-Pyoung Ahn, and Jae-Chul Lee Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.9b00765 • Publication Date (Web): 14 May 2019 Downloaded from http://pubs.acs.org on May 15, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26

Direct-contact micro-electrical measurement of the electrical resistivity of a solid electrolyte interface Jun-Hyoung Park1,†, Yong-Seok Choi1,†, Hyun-Jeong Lee1,2, Hyung-Cheoul Shim3, JaePyoung Ahn2, and Jae-Chul Lee1,* 1

Department of Materials Science and Engineering, Korea University, Seoul 02841, South Korea

2

Advanced Analysis Center, Korea Institute of Science and Technology, Seoul 02792, South Korea

3

Department of Applied Nano-Mechanics, Korea Institute of Machinery and Materials, Daejeon 34103,

South Korea

1020 1015

5 μm

SEI

1010 105

semi-conductors

1 10-5 10-10

insulators

Energy ( E-Ef, eV )

Resistivity ( Ω∙m )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

conductors

Ag

Ni

Ti

Ge

Si

4.0 2.0 0.0

3.7 eV

-2.0 -4.0

Γ XY Γ ZR

Γ

TU

Γ

V

glass TiO2 SiO2

Abstract: Because of its effectiveness in blocking electrons, solid electrolyte interface (SEI) suppresses decomposition reactions of electrolyte and contributes to the stability and reversibility of batteries. Despite the critical role of SEI in determining the properties of batteries, the electrical properties of SEI layers have never been measured directly. In this paper, we present the first experimental results of the electrical resistivity of a LiF-rich SEI layer measured using a direct-contact micro-electrical device mounted in an electron microscope. Measurements show that the SEI layer exhibits high electrical resistivity (2.3 × 105 Ωm), which is comparable with those of typical insulating materials. Furthermore, a combined technique of advanced analyses and first-principles calculations show that the SEI layer is mainly composed of amorphous LiF and a minute nanocrystalline Li2CO3 compound. The

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

electronic origin responsible for the high resistivity of the SEI layer is elucidated by calculating the band structures of various LixF compounds and interpreting their effects on the resistivity. This study explains why SEI can prevent the degradation of electrode materials and consumption of Li ions in the electrolyte, and thus can be viewed as a stepping stone for developing highly stable and reversible batteries.

Keywords Li-ion battery, Solid electrolyte interface, Electrical resistivity, Four-point-probe technique, First-principles calculations

†These authors contributed equally to this work. *To whom correspondence should be addressed: Jae-Chul Lee (email: [email protected])

ACS Paragon Plus Environment

Page 2 of 26

Page 3 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Repetitive in- and out-flows of cations through electrode surface during battery cycles promote reactions between electrolytes and electrodes, causing the formation of thin reaction layers (i.e., solid electrolyte interface, SEI) at the electrolyte/electrode interface1, 2. While conductive to carrier ions, this layer is resistant to electrons and thus, acts as a physical/electrochemical barrier that suppresses the further degradation of the electrode surface and consumption of Li ions in electrolyte3. Consequently, the development of batteries with high cyclic stability and reversibility relies on how well SEIs can inhibit the reaction between electrolytes and electrodes. In general, the ability of SEI layers to suppress reactions between electrolyte and electrode is determined by its electrical resistance (R); the higher the electrical resistance of SEIs, the more difficult it is to conduct electrons through the SEI layer. Moreover, the electrical resistance of SEI layers increases over repeated charge/discharge processes, because the thickness of SEI layers increases spontaneously over these processes (as per the relation 𝑅 = 𝜌 × 𝑡/𝐴, where A is the area, t the thickness, and ρ the electrical resistivity of SEIs). The thickening of the SEI layer would proceed until the electron conduction from electrodes to electrolytes is blocked. The electrical resistance at this stage is regarded as the critical/minimal resistance that can prevent the additional formation of SEIs and thus, further decomposition of electrolytes. This resistance value is particularly of interest in the field of surface coatings (i.e., artificial SEIs) of electrodes4, 5; if coated layers have an electrical resistance higher than the minimal resistance, they can suppress the decomposition of electrolytes and electrodes and thus, stabilize the electrode surface. Therefore, the measured electrical resistance of SEI layers can be used as a reference for designing coatings with high surface stability and thus, developing highly reversible batteries6, 7. Although SEIs were discovered and have been studied for four decades since 19791, only recently researchers have begun to recognize the all-important significance of their

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 26

electrical properties. The electrical resistance of SEIs has frequently been evaluated via cyclic voltammetry8 or electrochemical impedance spectroscopy (EIS)3, 9 performed on half or full cells. However, the values obtained from half- or full-cell experiments are affected by several variables such as the electrode itself10, additives in the electrolyte11,

12

, and residual Li

compounds13. Furthermore, the resistance deduced from EIS corresponds to the overall value reflecting both ionic and electronic contributions. More importantly, SEI is an interfacial layer with a thickness of several tens to several hundreds of nanometers and thus occupies only a very small volume (usually less than 0.3%) of the entire graphite electrode. Due to weak signals emitting from SEIs and the intricacy in interpreting electrical responses of EIS, previously introduced experimental methods based on half/full cells make it nearly unfeasible to solely evaluate the unique properties of SEIs among various effects collected from batteries. One way to unlock this ambiguity is to measure the electrical resistivity of SEIs directly. To directly measure the electrical properties of highly sensitive and ultrathin SEIs only, an experimental method that satisfies the following conditions is needed. First, all the preparation, processing, and analysis of the sample should be conducted in a confined space that limits the contact with oxygen and moisture. Second, analyses should be capable of detecting feeble signals from a thin and small SEI layer. A micro-electrical measurement based on a four-point-probe (FPP) device satisfies these experimental requirements if the system is mounted inside a scanning electron microscope (SEM) together with a vacuum transfer system. Using this setup, we directly measured the electrical resistance of few hundred-nm-thick SEI layers formed in Li-graphite half cells. The SEI layer observed in these cells exhibited a significantly high electrical resistivity (2.3 × 105 Ω∙m), which is comparable with those of typical insulating materials. A combined technique of advanced analyses and first-principles calculations revealed that SEI layers are mainly composed of an amorphous LiF compound along with a minute amount of nanocrystalline Li-carbonate (Li2CO3) particles. The electronic

ACS Paragon Plus Environment

Page 5 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

structures responsible for the high resistivity of the SEI layer were elucidated by calculating the band structures of various LixF compounds.

(b)

(a)

(c)

(d)

~250 nm

10 μm

(e)

5 μm

Pt-coating

1 μm

(f)

1 μm

(g)

Cr-coating

Pt (coating) Cr (coating)

SEI

F (SEI)

graphite

200 nm

5 nm

200 nm

C (graphite)

Figure 1. Secondary electron images of graphite particles recorded; (a) before and (b) after battery operation, showing the formation of SEIs on the surface of graphite particles. (c) Top and (d) side views of the SEI layer with a thickness of 250 nm. (e) Transmission electron microscopy (TEM) image of an SEI particle formed on the graphite surface. The selected area diffraction pattern (the inset image) shows that the SEI is composed of an amorphous phase with a small amount of nanocrystalline particles. (f) High-resolution TEM image recorded from (e), showing the presence of a small amount of nanocrystalline particles (indicated by the arrow). (g) Energy dispersive spectroscopy (EDS) elemental mapping corresponding to (e), showing the positions of C (blue), F (red), Cr (green), and Pt (yellow). Note that the elemental maps of Cr and Pt are from Cr- and Pt-coatings applied to the SEIs to prevent ion beam damage during sample preparation.

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Observation and chemical analyses of SEI layers. Before measuring the electrical resistance of the SEI layer, we observed the morphology and distribution of SEIs from the surface of electrode samples. Conditions employed in this experiment were strictly controlled such that SEI layers used for observations/measurements do not contain any unintended artifacts (including an artifact LiF compound) that could be formed by electron beam14, air15, and moisture16 (for details, see Figures S1-2 of Supporting Information). Figures 1a-b show the surface morphologies of graphite particles prepared before (Figure 1a) and after (Figure 1b) battery operation of the coin cell, respectively. The surfaces of graphite particles before battery cycles are devoid of any reaction products (see Figure 1a), whereas those after battery cycles (at a charging rate of 0.02C) are covered entirely by numerous SEI particles (see Figures 1bc). These SEIs are distributed uniformly on the surface of graphite particles and form a layer with a thickness of ~250 nm (Figure 1d). Analyses of the composition and crystallinity of SEI layer using transmission electron microscopy (TEM) equipped with EDS show that the SEI particle is a F-rich amorphous phase with a small amount of nanocrystalline particles (Figures 1e-g). As shown in the subsequent section, this amorphous phase is identified as a LiF compound, whereas nanocrystalline particles are considered to be Li-carbonate (Li2CO3) compounds17, 18. SEIs formed during the battery cycle are concentrated only at regions close to the electrode–electrolyte interface, whereas their number densities became markedly lower with distance away from the electrode–electrolyte interface (for details, see Supporting Information Figure S3). Because of the thin thickness of the SEI layer and its concentrated distribution on the surface, the SEI layer occupies only a very small volume (less than 0.3%) of the entire graphite electrode. Therefore, it is necessary to employ a plausible technique that can detect a weak signal, which otherwise is concealed when measured using half- or full-cell experiments.

ACS Paragon Plus Environment

Page 6 of 26

Page 7 of 26

Table 1. Results of the electrical resistivities measured for 250-nm-thick Ag, Ni, and Ti films. For comparison purpose, the electrical resistivities obtained in previous studies are also shown. Materials Ag Ni Ti

(a)

Measured resistivity ( × 10-8 Ωm) 1.28 6.00 39.6

I2

V2

Literature value ( × 10-8 Ωm) 1.59 [19] 6.99 [20] 42.0 [20]

(c) 1020

Range of resistivity This study

I1

V1

5 μm

(b) R

Resistivity ( Ω∙m )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1015

SEI

1010

Semi-conductors

Insulators

105 1 Conductors

10-5 I

10-10

Ag [19] Ni [20] Ti [20] Ge [21] Si [21] glass [21] TiO2 [21] SiO2 [21]

Figure 2. (a) The FPP test setup mounted inside an electron microscope, showing the SEI sample and four W-probes used for the resistivity measurement. (b) Changes in the resistance and current, measured as a function of applied voltage. It is worth noting that insignificant amounts of current are observed to flow when increasing the external potential from 0 to 5 V, a typical characteristic of dielectric materials. (c) The resistivity of the SEI layer, which is superimposed with the resistivities of typical materials in the literature.

Direct-contact electrical resistivity measurements. Before measuring the electrical resistivity of an SEI layer, we verified whether the FPP micro-electrical technique employed in this study could suitably measure the electrical resistivity of ultrathin materials with a thickness of ~250 nm. For this purpose, we prepared 250-nm-thick films of Ag, Ni, and Ti on

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

glassy SiO2 plates (see Supporting Information Figure S4), and measured their electrical resistance. Further, we converted the measured resistances to electrical resistivities by considering the geometry of the test samples (see Methods). As shown in Table 1, the resistivity values measured for Ag, Ni, and Ti films are similar to those reported in the literatures, indicating that the FPP measurement technique used in this study is suitable for measuring the resistivity of the ultrathin layer of SEIs. To measure the electrical resistivity of the SEI layer, we positioned a measurement setup similar to the one explained above inside an electronic microscope (see Figure 2a). When the current was measured by sweeping the voltage from 0 to 5 V in the absence of an electron beam, the SEI layer exhibits a very low and constant current value within the range of the applied voltage (see Figure 2b), which is typical of insulating materials. When the measured resistance was converted to the electrical resistivity, it was found to be 2.3 × 105 Ω·m. Furthermore, when this value was compared with those of several materials (see Figure 2c), it was found to be considerably higher than those of semiconducting materials and approaches those of insulators19-21. Electrically resistive SEI layer effectively blocks the passage of electrons and thus suppresses further decomposition reactions of the electrolyte with electrode surfaces. This reduces the further degradation of electrode materials and also prevents the consumption of Li ions in the electrolyte3 ; thus, improving the stability and reversibility of the batteries.

ACS Paragon Plus Environment

Page 8 of 26

Page 9 of 26

103

(b)

C 1s

C-C

2 6×10 8×102 2 4×10 6×102

2×1022 4×10

C=O C-O3

0 2×102 294

6×103 5×10

C-O

Counts /s

(a)108×10 1023

Counts /s

292

290

288

286

292

290

288

286

284

282

280

0

F 1s

Li-F

3

4×103 3×103 2×103

P-F

1×103

0

294

284

282

280

692

Binding energy ( eV )

(c)

690

688

686

684

682

680

Binding energy ( eV )

Li2CO3 [17] 5.1 eV

Intensity ( a.u. )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

LixF [17] 7.2 eV 5.1 eV

SEI

7.25 eV

60

70

80

90

100

Energy ( eV )

Figure 3. (a) C 1s X-ray photoelectron spectroscopy (XPS) spectra for the SEI layer. The yellow, orange, green, and blue lines represent deconvoluted peaks corresponding to C−O3, C=O, C−O, and C–C bonds, respectively. (b) F 1s XPS spectra for the SEI layer. The green and blue lines represent deconvoluted peaks corresponding to P–F and Li–F bonds, respectively. (c) Electron-energy-loss spectroscopy (EELS) spectra obtained from the SEI layer (Figure 1f) with Li-K edge, which is superimposed with those obtained from Li2 CO3 and LixF (with x ≈ 1). The inset is the magnified graph of the EELS spectra recorded from the SEI layer, indicating the presence of a small extra peak (indicated by the arrow) at 65.3 eV corresponding to the Li2CO3 compound.

Phase identification of SEI layers. In order to elucidate the structural origin of the high electrical resistivity of the SEI layer, it is necessary to identify the chemical compounds

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

comprising the SEI layer and their respective structures. Because C, F, and Li are the main elements of the electrolyte that react with the graphite to form SEIs, the bonding states of these constituent elements were first evaluated using XPS. C 1s and F 1s X-ray photoelectron spectrometry (XPS) spectra, due to their ability to analyze the chemical bonding states of the constituent elements of the SEIs, are useful to deduce the compounds constructing the SEI layer. Figures 3a-b show the narrow-scan XPS spectra for C 1s and F 1s, respectively. Upon the deconvolution of the C 1s spectra, four peaks are detected, each of which corresponds to the characteristic binding energy of C−O3 (290.0 eV), C=O (288.0 eV), C−O (286.3 eV), and C−C (284.7 eV). Among these peaks, the peak corresponding to the C−C bond is most likely to originate from the binder, carbon black, and various hydrocarbons used for the graphite electrode. In contrast, the peaks corresponding to C−O3, C=O, and C−O are considered to stem from the electrolyte decomposition compounds such as Li-carbonate (most likely Li2CO3 as evaluated from electron-energy-loss spectroscopy (EELS) in Figure 3c)17, Li-alkyl carbonates (R-OCO2Li), and Li-alkoxides (R-CH2OLi). A close examination of the curves deconvoluted from the F 1s XPS spectra (Figure 3b) shows that the F 1s spectra are separated into two characteristic peaks, each of which is located at 685 and 687.6 eV. The peak at 685 eV (in green) corresponds to Li−F bonds, whereas the lower intensity peak (in green) at 687.6 eV corresponds to P−F bonds, which probably emanates from a minor residue of the LiPF6 salt that remained after washing the anode materials22. It should be noted that F-O bonds are absent in the F 1s spectra. This indicates that SEIs observed in this study did not react with moisture and thus, do not contain any moisture-induced artifacts, which are typically revealed as POF3 compounds23. The absence of the moisture-induced artifacts was again confirmed by measuring the O 1s XPS spectra of SEIs (for details, see Supporting Information Figure S2). Based on the XPS results of C 1s, F 1s, and O 1s spectra presented above, it is apparent that the SEI layer consists of Li-rich compounds with Li-carbonate and Li−F bonds.

ACS Paragon Plus Environment

Page 10 of 26

Page 11 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Although the analyses based on XPS are instructive for deducing the presence of compounds with Li−F and Li-carbonate bonds, the results only disclose the chemical bonding states and thus are insufficient to identify the exact composition of the SEI layer. EELS spectra can provide a more detailed description of possible compounds, because they can evaluate the loss energy corresponding to the chemical bonding state of a chemical compound itself. Figure 3c shows the Li-K edge EELS spectra obtained from the SEI layer. For comparison purpose, the Li-K edge spectra corresponding to Li2CO3 and LixF compounds are also shown24. When comparing the three spectra, the characteristic loss energy (7.25 eV) obtained from the SEI layer coincides with that of the LixF. A careful examination of the spectra revealed the presence of a small additional peak (denoted by the arrow in the inset of Figure 3c). This corresponds to the Li2CO3, which is considered to reside as nanocrystalline particles observed in Figure 1e. These results indicate that LixF compounds are the major component comprising the SEI layer, whereas Li2 CO3 is minor. Although Li-bearing compounds such as R-OCO2Li, R-CH2OLi, and Li2CO3 have been commonly reported in both previous and present studies as the decomposition products of ethylene carbonate (EC) - dimethyl carbonate (DMC) electrolyte system17, 25, our results based on TEM, XPS, and EELS analyses confirms that the SEI layer consists of amorphous LixF phases (a-LixF) with a minor amount of nanocrystalline Li2 CO3.

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4. (a) Representative cooling curve of LixF1-x (Li0.25F0.75) used to prepare the amorphous LixF1-x phase obtained by quenching. (b) Potential energies of various LixF1-x phases evaluated using first-principles calculations. The images in the insets show the representative atomic structures of amorphous Li0.25F0.75, LiF, and Li0.75F0.25 phases prepared to evaluate the potential energy.

First-principles calculations. After identifying that the main component of the SEI layer is aLixF, the next task is to determine the composition of the LixF phase and identify its electronic structure. This is because the electrical resistivity has a close relationship with the electronic structure of a material/phase comprising the SEI layer. Considering that the SEI layer (i.e., aLixF1-x phase) was formed after a long time (50 h per each battery cycle), the composition of the a-LixF1-x phase is regarded as that of the stable a-LixF1-x phase and thus corresponds to the phase with a minimal internal energy. Therefore, the exact composition of the phase can be determined by calculating the internal energies of various a-LixF1-x phases. For this purpose, we first quenched seven different LixF1-x (where x takes the values 0.25, 0.41, 0.50, 0.52, 0.53, 0.59, and 0.75) to obtain the a-LixF1-x model phases by quenching them from 1500 to 300 K (for details of the calculation, see Figure 4a and Methods), and subsequently calculated the internal energies of the resultant a-LixF1-x phases. Figure 4b shows the changes in the internal energies calculated for the computationally generated a-LixF1-x phases. The results show that, among various a-LixF1-x phases, the a-LiF phase has the lowest internal energy and thus can be a plausible phase comprising the SEI layer. This composition is also consistent with that deduced from the Li–F binary equilibrium phase diagram.

ACS Paragon Plus Environment

Page 12 of 26

Page 13 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 5. Representative band structures calculated for amorphous (a) Li0.25F0.75, (b) LiF, (c) Li0.52 F0.48, and (d) Li0.75F0.25 phases. (e) Electrical resistivities calculated for the various LixF1x

phases. The error bars indicate the maximum and minimum value calculated for four to five

atomic structures per point.

To interpret the electronic origin of the high electrical resistivity observed from the SEI layer, we further investigated the electronic structures of the a-LixF1-x phases with differing compositions. Figures 5a–d show the examples of the band structures calculated for the a-LixF1x

phases with x being close to or away from 0.5 (e.g., x = 0.25, 0.50, 0.52, and 0.75). An

examination of the obtained results reveals that the band structures of the a-LixF1-x phases can be classified into two different types: those without and with a band gap. In the cases of aLi0.25 F0.75 and a-Li0.75F0.25 phases (see Figures 5a and d), the energy bands of these phases are concentrated near the Fermi energy (denoted by gray dashed lines), which is similar to typical conductive materials. Conversely, the band structures of the a-LiF and a-Li0.52F0.48 phases (Figures 5b and c) exhibit large band gaps of 3.7 and 3.2 eV, respectively, indicating that the electrical resistivity of these phases is comparable or close to those of insulating materials. The above analyses on the band structures of the a-LixF1-x phases suggest that the electronic

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

structures and corresponding electrical resistivity depend significantly on the Li composition of the a-LixF1-x phases. The composition dependence of the electrical resistivity (𝜌) can be roughly estimated 3⁄ 2 exp⁡(𝐸𝑔 /𝑘𝑇),

using 𝜌 = 𝜌o 𝑇 −

where T, k, and Eg are the temperature, Boltzmann constant, and

bandgap, respectively. Following this equation, the electrical resistivity of the a-LiF phase is higher than that of the a-Li0.52F0.48 phase by approximately eight orders of magnitude. However, this equation only considers the value of bandgaps of material in predicting the resistivity and thus cannot be applied to predict the resistivity of the a-LixF1-x phases with metallic band structures, such as the Li0.25F0.75 and Li0.75F0.25 phases. Therefore, to more precisely evaluate the composition dependency of the electrical resistivity, a more general approach to obtain the electrical resistivity of the a-LixF1-x phases is needed. A possible potential approach to achieving this is to solve the Boltzmann's equation26, which incorporates the entire band structures of each a-LixF1-x phase (see Figures 5a–d and Methods), and is explained in the following. Figure 5e shows the changes in the electrical resistivity of the a-LixF1-x phases evaluated as a function of composition (x). The results show that the electrical resistivity of the a-LixF1-x phases increases drastically as the x value approaches 0.5. We consider this rapid increase in the resistivity to originate from strong ionic bonds between Li and F ions. Under this tight bonding condition, only a few free electrons are shared by the ionic bond between Li and F ions27, which would consequently block the passage of electrons. Considering that the SEI layer is mainly composed of the a-LiF phase (see Figures 3 and 4), the above analyses on the electrical resistivity of the a-LixF1-x phases explain why the SEI layer is electrically resistive and thus can prevent the degradation of electrode materials and consumption of Li ions in the electrolyte. The present finding can serve as a basis for developing highly stable and reversible batteries. However, more systematic future studies on this topic are necessary.

ACS Paragon Plus Environment

Page 14 of 26

Page 15 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

In conclusion, using an FPP device mounted in an electron microscope, we performed a micro-electrical measurement of the electrical resistivity of the SEI layer formed uniformly on the surface of graphite particles in Li-graphite half cells. The SEI layer displays a high resistivity (2.3 × 105 Ω∙m) comparable with those of typical insulating materials. From advanced analyses and first-principles calculations, we elucidated that the SEI layer is mostly composed of the a-LiF phase with a minor amount of nanocrystalline Li2CO3 compound. Further calculations showed that the a-LiF phase is characterized by a bandgap of 3.7 eV, explaining why the SEI layer exhibits a high electrical resistivity. Although more systematic future studies on the changes in the electrical properties of SEIs (during repeated battery cycles) are necessary, our results can be used as a reference data for comparing the structural evolution and analyzing the respective properties of various SEIs in different battery systems. Furthermore, the present results can be viewed as a stepping stone for developing highly stable and reversible batteries by correlating the measured value with the electrical properties of intended a-SEIs. Finally, the methods of measurement presented in this paper on the electrical resistivity of SEIs provide a quantitative guideline for designing surface coatings (artificial SEIs) with high surface stability.

Methods SEI sample preparation. Lithium-ion batteries are the most widely explored batteries, whose properties are well-known28. Consequently, a Li-rich SEI is an excellent choice for evaluating its properties for the future development of SEIs with enhanced electrical performances. First, we synthesized a positive electrode material by mixing artificial carbon (MTI), polyvinylidene fluoride (PVdF; Solef 5130, Solvay), carbon black (Super P, d50 = 65 nm), and N-methyl-2pyrrolidinone (NMP; Carl Roth) using solid-state reaction. Then, we poured the powder

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

mixture into a Cu current collector and cast it into a 250-nm-thick strip using a knife coating technique; followed by drying at 750 °C for two hours in vacuum. Next, we assembled the synthesized electrode material (i.e., the cathode) and Li metal (i.e., the anode) in order to prepare 2032-type standard half cells. To prepare these cells, we utilized a monolayer polypropylene membrane (Celgard 2500, Celgard) as a separator, and used 1 M LiPF6 dissolved in 1:1 vol% mixture of EC and DMC as an electrolyte. Further, we performed cyclic voltammetry tests using a multichannel potentiostat (VMP3, BioLogic) in a constant-current mode by varying the potential from 0.05 to 2.50 V at a charging and discharging rate of 0.02 C. After the cyclic tests, we disassembled coin-cell samples to remove the graphite electrode, which was then washed in DMC solvent for 1 min to eliminate the electrolyte residue, followed by drying at 25 °C for 24 h in an air-tight Ar-flowing glove box. Subsequent SEM observations and XPS analyses performed on SEI samples revealed that this washing process/condition could effectively remove electrolyte residues from the surface of electrode samples (for details, see Supporting Information Figure S5). Furthermore, we moved the dried graphite electrode to a field-emission SEM (Quanta, FEI) to observe the morphologies of the SEI layer.

Direct-contact measurement of electrical resistivity. An appropriate method is required to measure the electrical resistivity of the SEI layer, which permits direct measurements of the electrical resistivity of an ultrathin SEI layer while diminishing the contact resistance. Furthermore, sample preparation and measurement must be performed under air-tight conditions because SEIs are vulnerable to oxygen and moisture. The FPP micro-electrical technique combined with electron microscopy and vacuum transfer system satisfies the abovestated experimental conditions (for details, see Supporting Information Figure S6). Before measuring the resistivity of the samples, four W tips were sharpened using a focused-ion beam device and positioned on the sample surface at 5-μm intervals. The FPP measurements were

ACS Paragon Plus Environment

Page 16 of 26

Page 17 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

then conducted by sweeping the bias voltage (from 0 to 5 V) between the two outer tips (V1 and V2), while the respective current (I) and voltage drop (V) between the two inner tips (I 1 and I2) were recorded (for a detailed test setup, see Figure 2b). Current-voltage (I-V) curves of the samples were recorded in the absence of an electron beam to avoid any potential beaminduced influence on the measurements. The measured resistance (R) of the SEI layer was converted to the resistivity (𝜌) using 𝜌 = 𝑅 × (𝑡 × 𝑤)/𝐿,

(1)

where t is the thickness, w is the width, and L is the length of specimens.

Microstructural and chemical analyses. To observe the microstructures and determine the chemistry of the SEI, the graphite electrode samples that were prepared from the one-cycled coin cells were transferred to the analytical devices using an air-tight transfer system to prevent the potential exposure to air. Electron diffraction patterns and elemental mappings of the selected area in the SEI region were collected using high-resolution TEM (Talos, FEI) equipped with an imaging filter spectrometer (Gatan-200, Gatan) operated at 200 kV. The chemical bonding states of each element comprising the SEI were assessed using XPS (PHI 5000 VersaProbe, Ulvac) by employing the Monochromatized Al Kα (1486.6 eV) radiation as the Xray source. The XPS data were calibrated by shifting all peaks relative to the lowest measured C 1s peak (284.6 eV). The main compounds that constituted the SEI were identified by evaluating EELS spectra in HR-TEM.

First-principles calculations. Car–Parrinello molecular dynamics simulations using a canonical ensemble were performed to determine the chemical composition of the SEI layer, and elucidate why it displayed a high electrical resistivity. To this end, seven different crystalline LixF1-x (c-LixF1-x) phases with x = 0.25, 0.41, 0.50, 0.52, 0.53, 0.59, and 0.75 were

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

first prepared as the initial structures. These samples were obtained by arranging 64 Li and F atoms to occupy the atomic positions corresponding to a (2 × 2 × 2) LiF supercell structure. The c-LixF1-x phases were then melted by heating to 1500 K followed by quenching to 300 K at 50 K/ps to obtain the a-LixF1-x phases that replicate the SEIs formed on the graphite electrode. All simulations were performed using ultrasoft pseudopotentials with Perdew–Burke– Ernzerhof exchange-correlation as implemented in the Quantum Espresso package with the temperature controlled by a Nosé–Hoover thermostat. The cutoff energy of 30 Ry was set for the plane wave expansion, and only the gamma point was sampled. The energy convergence was achieved down to 0.2 Ry/unit cell. The Verlet algorithm was used to solve the classical equations of motion. In order to select the plausible a-LixF1-x phase formed in the SEI, the potential energies of the seven different a-LixF1-x phases were then computed using the cutoff energy of 40 Ry and a Monkhorst–Pack 2 × 2 × 2 grid. To evaluate the electrical resistivities of the a-LixF1-x phases, the electronic structures of the seven a-LixF1-x phases were also calculated using the cutoff energy of 40 Ry and a Monkhorst–Pack 2 × 2 × 2 grid. This procedure ensured the estimation of reasonable values for the electrical resistivity. In order to analyze the effects of the computed electronic structures on the charge transport in the a-LixF1-x phases, the electrical resistivity corresponding to each electronic structure was subsequently evaluated using the Boltzmann transport theory, as implemented in the Boltzmann transport properties (BoltzTraP) code.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.####

ACS Paragon Plus Environment

Page 18 of 26

Page 19 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1. Validation that SEIs do not contain e-beam-induced artifact LiF (PDF) 2. Validation that SEIs do not contain moisture-induced artifact LiF (PDF) 3. Distribution of SEIs in the graphite electrode along the thickness direction (PDF) 4. Direct measurement of the electrical resistivities of various materials using fourpoint-probe technique (PDF) 5. Test of the washing conditions for removing electrolyte residues (PDF) 6. Validation of sealing performance (tightness) of the vacuum transfer equipment (PDF)

AUTHOR INFORMATION Corresponding Author E-mail: [email protected] (J.-C.L.)

ORCID Jun-Hyoung Park: 0000-0002-5391-4135 Yong-Seok Choi: 0000-0002-3737-2989 Hyun-Jeong Lee: Hyung-Cheoul Shim: 0000-0001-5275-8671 Jae-Pyoung Ahn: Jae-Chul Lee: 0000-0002-9294-2163

Author Contributions † J.-H.P. and Y.-S.C. contributed equally to this work.

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Notes The authors declare no competing financial interest.

Acknowledgment This work was supported by the Samsung Research Funding Center of Samsung Electronics under project No. SRFC-MA1602-04 and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST, NRF-2018R1A2B6003927).

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

Peled, E. J. Electrochem Soc 1979, 126, 2047-2051. Peled, E.; Menkin, S. J. Electrochem Soc 2017, 164, A1703-A1719. Lu, P.; Li, C.; Schneider, E. W.; Harris, S. J. J Phys Chem C 2014, 118, 896-903. Kim, M. S.; Ryu, J.-H.; Lim, Y. R.; Nah, I. W.; Lee, K.-R.; Archer, L. A.; Cho, W. I. Nat. Energy 2018, 3, 889. Liu, Y.; Lin, D.; Yuen, P. Y.; Liu, K.; Xie, J.; Dauskardt, R. H.; Cui, Y. Adv Mater 2017, 29, 1605531. Andersson, A.; Henningson, A.; Siegbahn, H.; Jansson, U.; Edström, K. J Power Sources 2003, 119, 522-527. An, S. J.; Li, J.; Daniel, C.; Mohanty, D.; Nagpure, S.; Wood, D. L. Carbon 2016, 105, 52-76. He, Y.-B.; Li, B.; Yang, Q.-H.; Du, H.; Kang, F.; Ling, G.-W.; Tang, Z.-Y. J. Solid State Electrochem 2011, 15, 1977-1985. Zhang, S.; Xu, K.; Jow, T. Electrochim Acta 2006, 51, 1636-1640. Aurbach, D.; Markovsky, B.; Salitra, G.; Markevich, E.; Talyossef, Y.; Koltypin, M.; Nazar, L.; Ellis, B.; Kovacheva, D. J Power Sources 2007, 165, 491-499. Kang, K. S.; Choi, S.; Song, J.; Woo, S.-G.; Jo, Y. N.; Choi, J.; Yim, T.; Yu, J.-S.; Kim, Y.-J. J Power Sources 2014, 253, 48-54. Aurbach, D.; Talyosef, Y.; Markovsky, B.; Markevich, E.; Zinigrad, E.; Asraf, L.; Gnanaraj, J. S.; Kim, H.-J. Electrochim Acta 2004, 50, 247-254. Cho, D.-H.; Jo, C.-H.; Cho, W.; Kim, Y.-J.; Yashiro, H.; Sun, Y.-K.; Myung, S.-T. J Electrochem Soc 2014, 161, A920-A926. Sina, M.; Alvarado, J.; Shobukawa, H.; Alexander, C.; Manichev, V.; Feldman, L.; Gustafsson, T.; Stevenson, K. J.; Meng, Y. S. Adv Mater Interfaces 2016, 3, 1600438. Edström, K.; Herstedt, M.; Abraham, D. P. J Power Sources 2006, 153, 380-384. Ravdel, B.; Abraham, K.; Gitzendanner, R.; DiCarlo, J.; Lucht, B.; Campion, C. J Power Sources 2003, 119, 805-810.

ACS Paragon Plus Environment

Page 20 of 26

Page 21 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

Dedryvere, R.; Laruelle, S.; Grugeon, S.; Poizot, P.; Gonbeau, D.; Tarascon, J.-M. Chem Mater 2004, 16, 1056-1061. Tian, B.; Światowska, J.; Maurice, V.; Zanna, S.; Seyeux, A.; Klein, L. H.; Marcus, P. J Phys Chem C 2013, 117, 21651-21661. Tsuda, Y.; Omoto, H.; Tanaka, K.; Ohsaki, H. Thin Solid Films 2006, 502, 223-227. Kemp, W.; Klemens, P.; White, G. Aust J Phys 1956, 9, 180-188. Giancoli, D. C.; Miller, I. A.; Puri, O. P.; Zober, P. J.; Zober, G. P., Physics: principles with applications. Pearson Education Upper Saddle River, NJ: 2005; Vol. 4. Chabal, Y. J.; Cho, K.; Hinkle, C. L.; Longo, R. C.; Santosh, K. Material Matters 2013, 8, 104-107. Kanevskii, L.; Dubasova, V. Russ J Electrochem 2005, 41, 1-16. Sina, M.; Thorpe, R.; Rangan, S.; Pereira, N.; Bartynski, R. A.; Amatucci, G. G.; Cosandey, F. J Phys Chem C 2015, 119, 9762-9773. Tasaki, K.; Goldberg, A.; Lian, J.-J.; Walker, M.; Timmons, A.; Harris, S. J. J Electrochem Soc 2009, 156, A1019-A1027. Ricci, F.; Chen, W.; Aydemir, U.; Snyder, G. J.; Rignanese, G.-M.; Jain, A.; Hautier, G. Sci Data 2017, 4, 170085. Avci, R.; Flynn, C. Phys Rev B 1979, 19, 5967. Etacheri, V.; Marom, R.; Elazari, R.; Salitra, G.; Aurbach, D. Energy Environ Sci 2011, 4, 3243-3262.

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Secondary electron images of graphite particles recorded; (a) before and (b) after battery operation, showing the formation of SEIs on the surface of graphite particles. (c) Top and (d) side views of the SEI layer with a thickness of 250 nm. (e) Transmission electron microscopy (TEM) image of an SEI particle formed on the graphite surface. The selected area diffraction pattern (the inset image) shows that the SEI is composed of an amorphous phase with a small amount of nanocrystalline particles. (f) High-resolution TEM image recorded from (e), showing the presence of a small amount of nanocrystalline particles (indicated by the arrow). (g) Energy dispersive spectroscopy (EDS) elemental mapping corresponding to (e), showing the positions of C (blue), F (red), Cr (green), and Pt (yellow). Note that the elemental maps of Cr and Pt are from Cr- and Pt-coatings applied to the SEIs to prevent ion beam damage during sample preparation. 234x131mm (150 x 150 DPI)

ACS Paragon Plus Environment

Page 22 of 26

Page 23 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

(a) The FPP test setup mounted inside an electron microscope, showing the SEI sample and four W-probes used for the resistivity measurement. (b) Changes in the resistance and current, measured as a function of applied voltage. It is worth noting that insignificant amounts of current are observed to flow when increasing the external potential from 0 to 5 V, a typical characteristic of dielectric materials. (c) The resistivity of the SEI layer, which is superimposed with the resistivities of typical materials in the literature. 249x114mm (150 x 150 DPI)

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(a) C 1s X-ray photoelectron spectroscopy (XPS) spectra for the SEI layer. The yellow, orange, green, and blue lines represent deconvoluted peaks corresponding to C–O3, C=O, C–O, and C–C bonds, respectively. (b) F 1s XPS spectra for the SEI layer. The green and blue lines represent deconvoluted peaks corresponding to P–F and Li–F bonds, respectively. (c) Electron-energy-loss spectroscopy (EELS) spectra obtained from the SEI layer (Figure 1f) with Li-K edge, which is superimposed with those obtained from Li2CO3 and LixF (with x ≈ 1). The inset is the magnified graph of the EELS spectra recorded from the SEI layer, indicating the presence of a small extra peak (indicated by the arrow) at 65.3 eV corresponding to the Li2CO3 compound. 256x200mm (150 x 150 DPI)

ACS Paragon Plus Environment

Page 24 of 26

Page 25 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

(a) Representative cooling curve of LixF1-x (Li0.25F0.75) used to prepare the amorphous LixF1-x phase obtained by quenching. (b) Potential energies of various LixF1-x phases evaluated using first-principles calculations. The images in the insets show the representative atomic structures of amorphous Li0.25F0.75, LiF, and Li0.75F0.25 phases prepared to evaluate the potential energy. 260x110mm (150 x 150 DPI)

ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Representative band structures calculated for amorphous (a) Li0.25F0.75, (b) LiF, (c) Li0.52F0.48, and (d) Li0.75F0.25 phases. (e) Electrical resistivities calculated for the various LixF1-x phases. The error bars indicate the maximum and minimum value calculated for four to five atomic structures per point. 188x82mm (150 x 150 DPI)

ACS Paragon Plus Environment

Page 26 of 26