A review of ¹⁷O isotopic labeling techniques for solid-state NMR structural studies of metal oxides in lithium-ion batteries

Abstract

Recent advances in utilizing ¹⁷O isotopic labeling methods for solid-state nuclear magnetic resonance (NMR) investigations of metal oxides for lithium-ion batteries have yielded extensive insights into their structural and dynamic details. Herein, we commence with a brief introduction to recent research on lithium-ion battery oxide materials studied using ¹⁷O solid-state NMR spectroscopy. Then we delve into a review of ¹⁷O isotopic labeling methods for tagging oxygen sites in both the bulk and surfaces of metal oxides. At last, the unresolved problems and the future research directions for advancing the ¹⁷O labeling technique are discussed.

Graphical abstract

1. Introduction

Metal oxides, renowned for their distinct physical and chemical attributes, find extensive applications across various fields [[1], [2], [3], [4], [5], [6]]. In contemporary lithium-ion battery industries, they serve roles as cathodes [7,8], anodes [9,10], and solid electrolytes [[11], [12], [13], [14]], contingent upon their electrochemical and structural characteristics. The dominant materials in the lithium-ion battery cathodes are well-ordered layered compounds like Li-(Ni, Mn, Co)O₂ (NMC) and Li(Ni, Co, Al)O₂ (NCA), mainly due to their high energy density and commendable cycling stability [8,15]. In the past two decades, many metal oxides, such as SnO₂ [9], TiO₂ [16], MnO₂ [17], and Fe₃O₄ [18], have been intensively studied as anode materials [19] owing to the high lithium storage capacity, excellent cycling performance [17,18]. Furthermore, a variety of lithium-containing metal oxides, such as Li₇La₃Zr₂O₁₂ [20,21] and Li₃Zr₂Si₂PO₁₂ [14], function as solid electrolytes due to their excellent ionic conductivity, compatibility with high-capacity electrodes, and strong mechanical, thermal, and electrochemical stability [14]. These materials are being explored for their unique properties which are pivotal for boosting the storage capacity and cycle life of lithium-ion batteries.

Solid-state NMR spectroscopy is a widely used method for elucidating the local structure of solids [22]. The ¹⁷O nucleus, the sole stable NMR-active oxygen isotope, provides profound insights into the structure and dynamics of oxygen-rich solids through ¹⁷O solid-state NMR [22,23]. Up to now, ¹⁷O solid-state NMR has evolved into a standard method for uncovering the relationship between structure and properties in metal oxides [24,25]. However, the natural abundance of ¹⁷O isotope is only 0.037%, necessitating isotopic enrichment for solid-state NMR studies. Although ¹⁷O isotopes, often obtained as ¹⁷O-enriched H₂O (abbreviated as H₂¹⁷O) or O₂ (denoted as ¹⁷O₂) with 10–90% ¹⁷O levels, are costly (approximately €2600 for 1 g of 90% ¹⁷O-enriched H₂O and €5900 for 1 L of 70% ¹⁷O-enriched O₂), efficient ¹⁷O isotopic labeling is typically performed to prepare metal oxides for ¹⁷O NMR analysis. For the two readily available ¹⁷O labeled sources mentioned above, the typical ¹⁷O enrichment level for the two labeling methods ranges from 10 to 75 atomic percent (at%) [26]. Selecting the enrichment level involves balancing cost against sensitivity [[26], [27], [28]]. It is noted [26] that 20% enrichment provides sufficient sensitivity for one-dimensional (1D) spectroscopy, while correlation measurements, especially two-dimensional (2D) experiments, benefit from higher enrichments (35%–45% or more [26]). After years of progress, various isotopic labeling methods have been developed, which fall into two primary categories: bulk and surface-selective labeling. This review outlines the ¹⁷O isotopic labeling methods in the solid-state NMR structural studies of metal oxides for lithium-ion battery research. We specifically highlight recent advances in bulk labeling techniques including high-temperature calcination with ¹⁷O₂ and the treatment with H₂¹⁷O, as well as surface-selective methods including moderate thermal treatment with H₂¹⁷O or ¹⁷O₂.

2. Bulk labeling

2.1. Thermal treatment with ¹⁷O₂

The bulk labeling methods rely on the thermal diffusion of ¹⁷O anions from the surfaces to the bulk of metal oxides. When the sample encounters an ¹⁷O₂-enriched atmosphere and undergoes high-temperature treatments (≥300 °C), ¹⁶O in the surface lattice can be exchanged with ¹⁷O in the gas phase, and the resulting ¹⁷O in the oxide surface can migrate into the interior of the sample.

In a typical ¹⁷O₂ bulk labeling procedure, the sample usually undergoes pretreatment at elevated temperatures under vacuum in order to remove the impurities adsorbed on the surfaces. Subsequently, the sample is exposed to an ¹⁷O₂ atmosphere at high temperature (typically ≥ 300 °C) for several hours/days to enable ¹⁷O bulk labeling. Note that high temperatures or long duration time can accelerate the ^16/17O exchange, which speeds up the labeling process. However, excessively high temperatures may cause unwanted phase transitions for the materials and the labeling temperature and duration time are often carefully chosen. Seymour and co-workers [29] enriched Li₂MnO₃ at 850 °C for 24 h within an atmosphere of ¹⁷O₂ at a pressure of 200 mbar. They successfully elucidated detailed structural information of Li₂MnO₃ cathode using the thermal treatment-based bulk ¹⁷O labeling method in combination with density functional theory (DFT) simulations. They observed five distinctive ¹⁷O resonances in the ¹⁷O NMR spectrum of Li₂MnO₃ (as illustrated in Fig. 1). Corroborated with DFT calculations, these shifts are attributed to the stacking faults, 4i and 8j sites in the Li₂MnO₃ structure.

Fig. 1.

¹⁷O NMR variable-offset cumulative spectroscopy (VOCS) spectrum of ¹⁷O-enriched Li₂MnO₃ with the two regions of ¹⁷O shifts at X (2100–2450 ppm) and Y (1600–1950 ppm). The graph shows five highlighted peaks with red fit lines. Numbers atop the peaks signify the ¹⁷O shifts in ppm. Bracketed numbers denote the proportion of various O environments in the isotropic region's total integral. Reprinted with permission from Ref. [29]. Copyright 2016, American Chemical Society.

This method is also used to study solid electrolyte materials. Zhu et al. [14] developed oxide-type solid electrolytes with high ionic conductivity, such as Li₃Zr₂Si₂PO₁₂ (LZSP), using a 'skeleton-retained cationic exchange' method. Using ¹⁷O NMR, electrochemical impedance spectroscopy, and DFT simulations, they demonstrated that LZSP possesses low-coordinated lithium ions, leading to exceptional ionic conductivity at room temperature. In this study, the optimal labeling temperature for LZSP (400 °C for a duration of 24 h ¹⁷O₂-based bulk enrichment) can be used for anode materials [22,30,31], too. For example, the ¹⁷O NMR spectrum of an anatase TiO₂ sample which is enriched with ¹⁷O₂ at 500 °C for 12 h, shows a clear peak at 558 ppm, indicating the tri-coordinated oxygen (O_3C) species in the bulk of this anode material [22].

2.2. The treatment of H₂¹⁷O

The treatment of H₂¹⁷O involved synthesizing with H₂¹⁷O and mechanical processing of ball milling (BM). John and co-workers [32] presented an ionothermal method for the simple and cost-effective enrichment of metal oxides with ¹⁷O. The study demonstrates that incorporating microlitre quantities of ¹⁷O-enriched water (∼50 μL of 35% H₂¹⁷O) in the ionothermal synthesis process leads to SIZ-4 (a model of AIPO zeolite with the chabazite topology) samples with sufficient ¹⁷O enrichment for high-quality ¹⁷O solid-state NMR spectra. This approach also enables an analysis of how ¹⁷O NMR parameters are influenced by the local structural environment, providing insights valuable for future studies of metal oxides. Suzi et al. [33] introduced a room-temperature ¹⁷O enrichment scheme of zeolites. This approach uncovers a dynamic and labile framework in zeolites, characterized by rapid and reversible bond breaking. They dehydrated zeolites at 300 °C under vacuum for 18 h and then mixed 50 mg of zeolites with 50 μL of H₂¹⁷O to create a slurry, which was sealed in a disposable PTFE (polytetrafluoroethylene) insert and placed inside a 4 mm ZrO₂ rotor for ¹⁷O NMR measurements. They proposed that the enrichment level to fall within the range of 15%–25%. The study shows that ¹⁷O enrichment occurs in both Si–O–Al and Si–O–Si linkages within the zeolite framework upon exposure to H₂¹⁷O, with Si–O–Al species enriching more rapidly. This enrichment is observed in various zeolite frameworks and is independent of the Brønsted acid proton, suggesting new opportunities for structural characterization of zeolites and metal oxides [33].

Mechanochemistry (such as BM) is a rapidly developing approach for synthesizing solid materials [[34], [35], [36], [37]]. By simply blending and grinding solid reactants, this method provides a swift and eco-friendly method of chemical synthesis [37]. In the way of BM with a small quantity of solvent to expedite the reaction, mechanical forces drive reactions that not only blend the reactants but also reduce particle sizes, thereby augmenting their reactivity on the surfaces. This liquid-assisted grinding technique presents cost-effective opportunities for labeling the bulk of oxides with ¹⁷O [[37], [38], [39], [40]]. Metro et al. [37] demonstrated a relevant scheme, wherein they subjected metal hydroxides and H₂¹⁷O precursor to BM, resulting in the production of ¹⁷O-labeled metal oxides with an enrichment level of ∼5%. In this method, H₂¹⁷O is added in quantities of less than 2 equivalents, allowing for isotopic modification of the hydroxyl group. The subsequent heat treatment step in an inert atmosphere transforms the labeled intermediates into the desired metal oxides, as depicted in Scheme 1 [37].

Scheme 1.

¹⁷O labeling scheme for the bulk of metal oxides via BM. Reprinted with permission from Ref. [37]. Copyright 2017, Wiley-VCH.

The same group [37] extended the BM-based bulk labeling method to 11 more samples, encompassing various materials, such as metal hydroxides (e.g. Mg(OH)₂), carboxylic acids (e.g. 1,1′-carbonyl-diimidazole), boron compounds (e.g. B(OH)₃) and crystal phases with coordination water ligands (e.g. Sr(OH)₂·8H₂O). These samples were prepared with approximately 220 μL of H₂¹⁷O during BM, taking less than 2 h for each synthesis. These samples exhibited average enrichment levels ranging from 3% to 11%. Chen et al. [38] modified the BM enrichment process, utilizing Li₂O, CaO, Al₂O₃, SiO₂, TiO₂, and ZrO₂ as initial materials to attain higher levels of ¹⁷O-bulk labeled hydroxides and heat-treated oxides (the maximum enrichment level is up to 32%). Rainer et al. [40] successfully applied this method to zeolites, achieving mechanochemical hydrolysis of germanium silicate zeolites at room temperature in just 30 min with a solvent ratio of 50 mL/g (liquid/zeolite), with the maximum level of enrichment being ∼11%. In lithium-ion battery materials research, bulk labeling is more commonly employed since the ion intercalation and deintercalation mainly take place in the bulk during battery charging and discharging. As a result, researchers often concentrate on analyzing the bulk structure and prefer to label the bulk [29,41].

3. Surface labeling

Using ¹⁷O surface-selective labeling, researchers have successfully realized the ¹⁷O NMR observations of metal oxide surfaces. Initially, this approach was primarily employed for the investigation of catalytic oxides. However, it may also find applications in the study of metal oxides for lithium-ion battery in the future, such as probing the electrode/electrolyte interfaces. At present, the ¹⁷O NMR research on relevant field predominantly stays at the surface structure of individual metal oxides, as elucidated below.

In general, the attainment of surface-selective labeling with ¹⁷O involves temperature regulation, which is contingent upon the choice of ¹⁷O source (H₂¹⁷O or ¹⁷O₂). When using ¹⁷O₂ as the source, it typically demands a relatively high temperature due to the low activity of O₂ at low temperatures, exemplified by materials like MgO (e.g., 350 °C is used) [42]. When opting H₂¹⁷O, lower temperatures prove adequate for surface labeling, a feature prominently observed in materials such as CeO₂ [43] and TiO₂ [22]. The exposure to H₂¹⁷O vapor or ¹⁷O₂ causes an exchange between the ¹⁶O anions in the surfaces of metal oxides and the ¹⁷O in H₂¹⁷O or ¹⁷O₂. To ensure the successful detection of surface oxygen using ¹⁷O NMR, two critical factors must be considered (taking H₂¹⁷O as an example): 1) To observe surface oxygen species, it's important to utilize materials characterized by small particle sizes and large specific surface areas (Fig. 2a) [24]; 2) The labeling temperature also plays a crucial role. It should strike a delicate balance, avoiding excessive heating that could cause the migration of ¹⁷O from the surface to the bulk phase, while still facilitating effective exchange between ¹⁷O and ¹⁶O (Fig. 2b) [23].

Fig. 2.

Two pivotal factors that facilitate successful ¹⁷O NMR studies of nano-sized metal oxides. Spheres marked with light blue and red colors respectively denote ¹⁶O and ¹⁷O atoms. Reprinted with permission from Ref. [24]. Copyright 2023, American Chemical Society.

Wang et al. pioneered an ¹⁷O surface-selective labeling scheme [23] using 2 nm ceria nanoparticles and H₂¹⁷O to enrich the ceria surfaces at 373 K. By adsorbing H₂¹⁷O onto ceria surfaces at room temperature, they distinctly labeled oxygen ions in the first and second layers, as well as hydroxyl groups at 1040 ppm, 920 ppm, and 270 ppm, respectively, excluding the observations of the internal third layer and bulk oxygen sites (see Fig. 3). This work highlighted the possibility of efficient surface-selective ¹⁷O isotope labeling at moderate temperatures, thereby advancing surface observations in ¹⁷O NMR spectroscopy of metal oxide. It serves as a model system for the development of surface labeling techniques.

Fig. 3.

(a) ¹⁷O solid-state NMR spectra of ceria nanocrystals enriched at different temperatures and acquired at different external fields with accompanying DFT-simulated NMR shifts. (111) surface is employed, and the calculated ¹⁷O shifts for each layer are presented on the right. (b) ¹⁷O MAS NMR spectra of ceria nanorods labeled by heating in ¹⁷O₂ at 923 K and adsorbing H₂¹⁷O at room temperature followed by thermal treatment at 373 K under vacuum. Reprinted with permission from Ref. [23]. Copyright 2015, American Association for the Advancement of Science.

Li et al. conducted an investigation to distinguish between various surfaces of titania [22]. Using the ¹⁷O surface-selective enrichment technique, they selectively labeled anatase titania nano-octahedra and nanosheets which were characterized by high-energy (001) and low-energy (101) facets, respectively. Notable differences were observed in their ¹⁷O NMR spectra. Resonances in the 600–800 ppm range were linked to bi-coordinated oxygen (O_2C) on the surfaces, while peaks in the 450–560 ppm range were associated with surface/subsurface O_3C. Importantly, the intensity of O_2C species exceeded that of O_3C, demonstrating the effectiveness of surface-selective labeling (refer to Fig. 4). This study presents strong evidence that when anatase titania nanosheets mainly exposing (001) surfaces come into contact with small amounts of water, surface reconstruction occurs. This leads to the dissociation of water and its subsequent adsorption. Furthermore, the researchers pinpointed the 'step edge' as the main defect on the (101) surfaces of anatase titania. The presence of oxygen atoms at these step edges significantly amplified peak intensities, suggesting its high reactivity. As a result, water molecules showed a preference to adsorb onto the (101) surfaces. This research highlights the capabilities of ¹⁷O NMR spectroscopy in decoding complex surface interactions and disclosing unique surface characteristics of different titania facets.

Fig. 4.

¹⁷O NMR spectra of anatase titania nano-octahedra and nanosheets respectively mainly exposing (101) and (001) surfaces (abbreviated as NS001–TiO₂ and NO101–TiO₂, respectively) in comparison with the spectrum of non-faceted (NF1–TiO₂) titania. ¹⁷O NMR resonances attributing to hydroxyl groups (OH), molecularly bonded water molecules (H₂O), and O_2C and O_3C sites are marked. Reprinted with permission from Ref. [22]. Copyright 2017, Springer Nature.

Afterwards, Chen et al. combined ¹⁷O surface labeling and DFT calculations to clarify a polarity compensation mechanism seen in the highly reactive ceria (100) surfaces [44]. This mechanism featured a pivotal interplay between CeO₄ terminated (CeO₄-t) reconstruction and surface hydroxyl groups. Through quantitative ¹H NMR analysis, they determined the relative concentrations of the dominant O-terminal (O-t) surfaces (57%) and the reconstructed CeO₄-t surfaces (43%) in the ceria (100) polar surfaces. In recent years, there have been significant advancements in ¹⁷O surface-selective labeling techniques. Beyond ceria and titania, these methods have been broadened to include a range of metal oxides such as ZnO [45,46], ZrO₂ [47], SnO₂ [48], Ta₂O₅ [49], signifying a rapidly expanding field of research in this area.

As lithium-ion batteries involve crucial interfacial interactions between electrodes and electrolytes, the ¹⁷O surface-selective-labeling techniques should not only find applications in the study of metal oxide structures as mentioned earlier, but also hold significant promise in enhancing our comprehension of the properties of metal oxides used in lithium-ion battery, electrode reaction mechanisms, and the dynamics occurring at the interfaces between electrode and electrolyte. It is anticipated that the forthcoming research will yield insights in this area.

4. Conclusion and perspective

¹⁷O solid-state NMR spectroscopy coupled with isotopic labeling can provide highly sensitive insights into the structure-property correlations of the metal oxides used in lithium-ion battery researches. Several ¹⁷O labeling schemes have been developed for the specific sites of materials. To probe bulk oxygen species, high-temperature calcination with ¹⁷O₂ and ball-milling with H₂¹⁷O isotopes are demonstrated. For surface analysis, low-temperature treatments using H₂¹⁷O or ¹⁷O₂ were proposed. The NMR intensities obtained with these methods rely on the ¹⁶O/¹⁷O substitution rate and the reactivity of the oxygen species being labeled. These labeling methods outlined in this review provide significant enhancement to the detection limit for the bulk and surface of metal oxide materials, allowing many oxygen sites of specific structure detectable in NMR spectroscopy. Given the high cost of ¹⁷O-labeled isotopes, it is imperative to explore more cost-effective and controlled approaches for future research. One such avenue is the utilization of atomic layer deposition (ALD) methods, using H₂¹⁷O as a precursor. This approach enables the controlled fabrication of surfaces, subsurfaces, and bulk materials, distinctly labeled with ¹⁷O, through different cycles. Additionally, external field assisted surface-selective labeling techniques such as plasma using ¹⁷O isotopes can be another high-efficiency scheme avoiding high-temperature treatments. Advancements in labeling methodology can significantly enhance sensitivity, enabling its broader application in various metal oxide materials used in lithium-ion batteries.

CRediT authorship contribution statement

Xiaoli Xia: Writing – original draft, Writing – review & editing, Methodology. Lei Zhu: Writing – original draft, Writing – review & editing. Weiping Tang: Writing – review & editing. Luming Peng: Writing – original draft, Writing – review & editing, Funding acquisition, Conceptualization. Junchao Chen: Writing – original draft, Writing – review & editing, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Luming Peng is an editorial board member of MRL but was not involved in the editorial review or the decision to publish this article.

Biographies

Xiaoli Xia obtained her B.S. in Chemical Engineering and Technology from Shanxi Datong University in 2016 and her M.S. in Applied Chemistry from Shanxi University in 2019. She is now pursuing her doctoral research on metal oxide surfaces at Nanjing University under the supervision of Prof. Dr. Luming Peng, using solid-state NMR techniques.

Lei Zhu earned her B.S. in Applied Chemistry from Nanjing University of Aeronautics and Astronautics in 2013 and her M.S. from the Shanghai Academy of Spaceflight Technology in 2016. She currently works at the State Key Laboratory of Space Power Sources, Shanghai Institute of Space Power-Sources, and is pursuing her Ph.D. in Physical Chemistry at Fudan University, mentored by Prof. Dr. Yongyao Xia. She is focused on researching oxide-type solid electrolytes and associated all-solid-state battery technologies.

Weiping Tang, Ph.D. in Chemical Engineering, currently serves as a full professor at Shanghai Jiao Tong University and is the Dean of the Shanghai Institute for Advanced New Energy Source Technologies. He is a committee member of the Physical Chemistry Power Source Section of the China Electronics Society, a member of the National Solid-State Ionics Association, and a board member of the Solid-State Battery Industry Alliance. Professor Tang has been engaged in the foundational and applied research of lithium batteries for over 20 years, with a primary focus on the application of solid-state oxide materials in lithium-ion adsorption materials and solid-state lithium batteries. He has published over 200 papers, authored 4 books (including co-authored and translations), and applied for over 60 patents.

Luming Peng received his B.S. in chemistry from Nanjing University in 2001, and Ph.D. from State University of New York at Stony Brook in 2006, working under the guidance of Professor Clare P. Grey. He did postdoctoral studies with Professor Jonathan F. Stebbins at Stanford University during 2006–2008. In 2008, Dr. Peng joined the faculty of School of Chemistry and Chemical Engineering at Nanjing University.

Junchao Chen earned his bachelor's degree in Applied Chemistry from Nanjing University of Aeronautics and Astronautics in 2013. He then pursued his Master's and Ph.D. in Chemistry from Nanjing University, completing them in 2015 and 2019 respectively, under the guidance of Prof. Dr. Luming Peng. Following his doctorate, he conducted postdoctoral research at Shanghai Jiao Tong University (2021–2022) and then at Radboud University (2021–2023). In August 2023, he became an Assistant Research Professor at Shanghai Jiao Tong University. His primary expertise and interest are in utilizing solid-state NMR and diffraction methods to address challenges in catalysis and solid-state batteries. Currently, his research centers on oxide solid electrolytes and all-solid-state battery techniques.