Facet-dependent Rock-salt Reconstruction on the Surface of Layered

Jan 2, 2018 - Facet-dependent Rock-salt Reconstruction on the Surface of Layered Oxide Cathodes. Hanlei Zhang, Brian M. May, Jon Serrano-Sevillano, Mo...
0 downloads 4 Views 5MB Size
Download PDF
Article Cite This: Chem. Mater. 2018, 30, 692−699

pubs.acs.org/cm

Facet-Dependent Rock-Salt Reconstruction on the Surface of Layered Oxide Cathodes Hanlei Zhang,†,‡ Brian M. May,‡,§ Jon Serrano-Sevillano,∥ Montse Casas-Cabanas,∥ Jordi Cabana,‡,§ Chongmin Wang,*,⊥ and Guangwen Zhou*,†,‡ †

Materials Science and Engineering Program & Department of Mechanical Engineering, State University of New York, Binghamton, New York 13902, United States ‡ NorthEast Center for Chemical Energy Storage, State University of New York, Binghamton, New York 13902, United States § Department of Chemistry, University of Illinois at Chicago, Chicago, Illinois 60607, United States ∥ CIC Energigune, Albert Einstein 48, Parque Tecnológico de Á lava, Miñano 01510, Spain ⊥ Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, Washington 99352, United States S Supporting Information *

ABSTRACT: The surface configuration of pristine layered oxide cathode particles for Li-ion batteries significantly affects the electrochemical behavior, which is generally considered to be a thin rock-salt layer in the surface. Unfortunately, aside from its thin nature and spatial location on the surface, the true structural nature of this surface rocksalt layer remains largely unknown, creating the need to understand its configuration and the underlying mechanisms of formation. Using scanning transmission electron microscopy, we have found a correlation between the surface rock-salt formation and the crystal facets on pristine LiNi0.80Co0.15Al0.05O2 primary particles. It is found that the originally (014̅) and (003) surfaces of the layered phase result in two kinds of rock-salt reconstructions: the (002) and (111) rock-salt surfaces, respectively. Stepped surface configurations are generated for both reconstructions. The (002) configuration is relatively flat with monatomic steps while the (111) configuration shows significant surface roughening. Both reconstructions reduce the ionic and electronic conductivity of the cathode, leading to a reduced electrochemical performance.

■

INTRODUCTION Layered oxide cathode materials for lithium ion batteries attract wide attention due to their outstanding capacity and rate capability.1,2 However, their application is limited by their structural instability,3,4 which results in reduced long-term electrochemical performance.5−8 This instability is largely attributed to the layered → rock-salt phase transformation occurring in the surface of the primary particles, resulting from its high chemical reactivity toward the environment.7,9 Even the surface of the pristine particles without electrochemical cycling is easily degraded to a rock-salt layer with a thickness of a few atomic layers.10−12 This surface rock-salt layer protects the bulk of the particle from further degradation,9,13 but it also exhibits a reduced ionic and electronic conductivity, which compromises lithium transport into the primary particles.5,11,14 This surface rock-salt layer on the pristine particles also serves as the nucleus for a continued layered → rock-salt phase transformation during the electrochemical cycling.15−19 Therefore, an in-depth understanding of the morphology and atomic configuration of the rock-salt layer on the surface of pristine particles is essential for comprehending the surface degradation and the associated fade of electrochemical performance as a function of cycling. There is currently a great lack of knowledge regarding the configuration and formation process of the surface rock-salt © 2018 American Chemical Society

layer on the pristine particles, except for its thin nature and spatial location on the surface.11,20 The rock-salt layer is simply assumed to uniformly distribute on the surface.13,15,21 Thus, the layered → rock-salt phase transformation is considered a simple diffusional process,22 in which the detailed structure and morphology of the surface rock-salt layer are not taken into consideration.13 However, here we show that the surface rocksalt growth formation depends on the crystalline facets of the primary particles of LiNi0.80Co0.15Al0.05O2 (NCA), rather than being a random process. With probe-aberration-corrected scanning transmission electron microscopy (STEM) techniques and thin-plate NCA particles, the facet-dependent rock-salt formation is identified at the atomic level. The structural anisotropy of the layered phase leads to the anisotropic growth of the surface rock-salt layer: along the orientations in the (003) plane of the layered phase, the diffusional and structural properties are sufficiently different compared with those along the c-axis, which is perpendicular to the (003) plane. Accordingly, the surface configuration of the rock-salt layer is found dependent on the crystal facets of the primary particles. Received: September 14, 2017 Revised: January 2, 2018 Published: January 2, 2018 692

DOI: 10.1021/acs.chemmater.7b03901 Chem. Mater. 2018, 30, 692−699

Article

Chemistry of Materials

Figure 1. (a) Atomic STEM-HAADF view of a pristine primary particle, which possesses the R3̅m layered structure with a thin layer of the Fm3̅m rock-salt phase formed on the surface (∼5 atomic layers thick). (b) Magnified view of the layered phase in (a). (c) Schematic showing the stacking of TM and lithium slabs along the [003] direction. (d) Low-magnification STEM image of a representative pristine primary particle showing the (003)type and non-(003) surfaces. (e) Schematic showing the (003) and non-(003) surfaces of a representative primary particle. In the degrading environment, the (003) and non-(003) facets develop differently configurated rock-salt reconstructions due to the structural anisotropy of the layered phase. were performed using the following microscopes, respectively: (1) FEI Titan 80-300 microscope equipped with a field emission gun (FEG) and a probe aberration corrector, operated at an acceleration voltage of 300 kV and (2) FEI Titan 80-300 microscope with an FEG and an image aberration corrector, operated at an acceleration voltage of 300 kV. STEM Simulation. STEM-HAADF simulation in this work was performed using the JEMS TEM simulation software. The microscopic parameters used for the simulation were taken from the Titan 80-300 microscope that was used for the experimental measurements.

The influence of the facet-dependent surface reconstructions of the rock-salt phase on the electrochemical performance of cathode is thus discussed.

■

EXPERIMENTAL SECTION

Sample Preparation. LiNi0.80Co0.15Al0.05O2 platelets were synthesized in a two-step process for observation. Ni(NO3)2·6H2O, Co(NO3)2·6H2O, and Al(NO3)3·9H2O were first mixed in water using the appropriate stoichiometric ratio, followed by addition of NaOH using a molar ratio of 1:2, in a total of 70 mL of solution. The mixture was hydrothermally treated at 160 °C for 12 h, using water in a Teflon-lined stainless vessel, producing the hydroxide precursor. The precursor was washed, dried, and then annealed at 400 °C for 4 h to decompose the layered double hydroxide structure.23 The resulting oxide was mixed with LiOH·H2O and subsequently calcined at 800 °C for 1 h, followed by 700 °C for 12 h. STEM imaging, energy dispersive spectroscopy (EDS), and electron energy loss spectroscopy (EELS) were obtained from the assynthesized material (Figure S1a−c), which are in good agreement with those of stoichiometric LiNi0.80Co0.15Al0.05O2 cathodes used industrially.13,24 High-resolution X-ray diffraction (XRD) data (Figure S1d) were collected using the 11 BM beamline at the Advanced Photon Source of Argonne National Laboratory in transmission geometry at 295 K with a calibrated wavelength of λ = 0.414540 Å, a step size of 0.001°, and a time per scan of 0.1 s, which yields results that are in good agreement with those of the NCA material with the 25 ideal R3m ̅ layered structure. Rietveld refinement of the highresolution XRD pattern shows that the as-synthesized material has parameters a = b = 2.86627(2) Å, c = 14.1909(1) Å, and c/a = 4.95, as well as occ Ni3a = 0.027(3) with χ2 = 4.117. Li/Ni exchange gives a compound with the following formula unit: [Li0.973Ni0.027]3aNi0.773Li0.027Co0.15Al0.05]3bO2. The Rietveld refinement results are also in good agreement with LiNi0.80Co0.15Al0.05O2 materials with an ideal R3̅m layered structure and stoichiometry.26,27 Details of the Rietveld refinement can be found in Figure S1e and Table S2. Transmission Electron Microscopy (TEM) Imaging. The pristine material was dispersed on a copper grid coated with a lacey carbon film for scanning transmission electron microscopy (STEM) high angle annular dark field (HAADF) and high-resolution transmission electron microscopy (HRTEM) observations, which

■

RESULTS Macroscale Morphology of Pristine Primary Particles. STEM-HAADF analysis confirmed that the pristine primary particle shows a layered structure (space group R3m ̅ ), covered by a thin surface layer of the rock-salt phase (space group Fm3̅m) with a thickness of ∼5 atomic planes (Figure 1a and Figure S2). HRTEM observation of the same material (Figure S3) confirms the presence of the surface rock-salt layer, rather than artifacts generated by the STEM observation. This rocksalt layer is a minor structural degradation occurring on the particle surface upon the synthesis process, likely due to reduction of the layered phase upon treatment at high temperature. A magnified view of the layered phase (Figure 1b) shows that it is alternating TM and lithium slabs stacking along the [003] direction, as schematically shown in Figure 1c. Due to the stacking of (003) slabs, the (003) surface of the layered phase is well shaped and straight. Differently, along the [120] direction shown in Figure 1c, the crystal configuration is more complicated and unlikely to form a well-defined straight surface; a surface with a relatively random shape would form instead. Due to the structural anisotropy of the layered structure (Figure 1b,c), the pristine primary particle develops two very different kinds of macroscale facets: one pair of long and straight facets that are parallel to the (003) plane of the layered lattice, different from the set of facets that are nonparallel to the (003) plane of the layered phase, as experimentally and 693

DOI: 10.1021/acs.chemmater.7b03901 Chem. Mater. 2018, 30, 692−699

Article

Chemistry of Materials schematically shown in Figure 1d,e. Surface reconstructions of these two kinds of facets are analyzed, as shown in the following. Atomic-Level Morphologies of Rock-Salt Reconstructions on the Non-(003) and (003) Surfaces. Figure 2a is a

Figure 3. (a) STEM view of the rock-salt reconstruction on a (003)type surface with pyramid-like structures. (b) Schematic showing the atomic configuration of the pyramid-like structure. (c) Zoom-out STEM image showing the rough surface resulting from the (111) rocksalt reconstruction. (d) Further zoom-out HRTEM view clearly showing the hill-and-valley surface of the (111) rock-salt reconstruction. (e) Magnified HRTEM view of a protruded area (hill) on the rough (111) reconstruction surface shown in (d).

Figure 2. (a) STEM-HAADF view of a reconstructed (002) rock-salt facet containing a sequence of monatomic steps. (b) Zoom-in view showing the (002) stepped surface of the rock-salt layer with the step riser being the (111) plane. (c) Schematic showing the crystallographic features of the stepped surface. (d) Crystallographic configuration of the stepped surface on the (002) rock-salt layer (showing the TM cations only), with the surface TM cations marked in purple, the bulk cations marked in gray, and the missing cations indicated by dashed circles.

smallest terrace width, and the layers beneath progressively develop into wider terraces, as schematically illustrated in Figure 3b. The terraces are (111) planes of the rock-salt phase and are separated by (002) microfacets. The high concentration of the pyramid-like structures transforms the surface into a hilland-valley structure, as shown in Figure 3d. A magnified HRTEM view of a peak structure (Figure 3e) clearly shows its Fm3̅m nature of the rock-salt phase. As shown in Figure 3d, some regions in the (111) reconstruction show an amorphouslike structure, which can be induced by lattice distortion and even a loss of crystallinity,15,20,32 indicating that the (111) reconstruction is much more defective compared with the (002) reconstruction. According to the observations above, rock-salt surfaces, with (002) and (111) orientations, are dominantly present on the primary particles (Figures 2 and 3). The atomic configurations of the exposed facets are largely dependent on their electrostatic properties, which determine whether their presence is energetically favorable. To simplify the discussion, here we assume that the rock-salt phase in the current work has the formula of NiO, although, in reality, it contains a small amount of Li, Co, and Al cations as well.33,34 In this model rock-salt system, the (002)-type surface is composed of 50% of Ni2+ cations and 50% of O2− anions, as schematically shown in Figure 4a. The equal number of cations and anions on the (002) surface leads to its electrostatic neutrality, suggesting that its exposure is energetically favorable,35 as observed in Figure 2. Conversely, the (111)-type planes of the rock-salt phase are composed of alternating anion/cation sheets stacking along the [111] direction (Figure 4b), so the exposed (111) facet is either positively or negatively charged, increasing the electrostatic energy and thus energetically unfavorable. Figure 4c schematically shows the electrostatically positive (111) surface with the Ni2+ cations on top (marked in purple). Therefore, the (111) surface tends to reconstruct to form the so-called “octopolar” structures on the (111) plane,36 generating an electrostatically neutral and stable (111) surface.

STEM view showing the rock-salt reconstruction on a non(003) facet, where a rock-salt layer of ∼5 atomic planes is present. The surface reconstruction features relatively long and straight (002) rock-salt facets (as identified by the diffractogram of Figure 2a in Figure S4) with atomic-height steps, as marked with orange dashed lines in Figure 2a. A zoom-in view of the stepped region (Figure 2b) confirms that the steps have a height of 1 atomic plane, with the riser between two steps being a (111)-type facet (Figure 2b,c). Previous work has shown that the rock-salt phase can form in the surface region via inward diffusion of TM cations along lithium channels,10,28 and Figure 2 indicates that this inward diffusion leads to (002) rock-salt reconstruction. Crystallographically, the stepped structure forms via missing half of the (002) atomic plane on the surface, as schematically shown in Figure 2d (the missing TM cations are represented by dashed circles). Since the (002)-type plane is the cleavage plane of the rock-salt structure,29−31 the formation of the stepped surface by exposing the (002)-type planes of the rock-salt lattice is both energetically and structurally favorable. Unlike the relatively flat rock-salt reconstruction on the non(003) surface shown in Figure 2, the reconstruction on the (003)-type facet of the layered particle has a much rougher morphology, as shown in Figure 3. The magnified STEM view in Figure 3a shows a pyramid-like (stepped) structure of the rock-salt reconstruction, which is more clearly shown in its zoom-out view (Figure 3c, marked with a dashed line). Compared with the (002) reconstruction in Figure 2, this reconstruction leads to significant roughness with a high concentration of monatomic steps making up pyramid-like structures (Figure 3a,c). The topmost atomic layer has the 694

DOI: 10.1021/acs.chemmater.7b03901 Chem. Mater. 2018, 30, 692−699

Article

Chemistry of Materials

Figure 4. (a) Projection view of the neutral (002) surface of the NiO rock-salt structure, which is composed of 50% of cations and 50% of anions. (b) 3-D view of the rock-salt structure showing the alternating stacking of the (111) anion/cation planes. In (a) and (b) the oxygen anions are represented by smaller red circles to emphasize the TM cations. (c) Projection view of the cation arrangement on the (111) surface of the rock-salt phase before reconstruction, where the surface cations are highlighted in purple with the bulk cations marked in gray. (d) Schematic showing the reconstructed (111) surface cations with an ordered arrangement, as marked in purple. (e) Schematic showing the reconstructed (111) surface cations with a disordered arrangement, as marked in purple. (f) Schematic showing side view ([11̅0] direction) of the loose, disordered (111) surface cations induced by the surface reconstruction, as shown in part e.

The electrostatic polarity of the (111) surface is eliminated by the absence of half of the Ni2+ cations on the surface, with the oxygen anions underneath thereby becoming surface anions. After this reconstruction, the surface is composed of 50% of cations and 50% of anions, which is electrostatically neutral. A surface cation and its neighboring anions compose an octopolar structure, as described in the literature.36 If the removal of the surface cations follows an ordered fashion that does not change the lattice symmetry of the (111) plane, a (111)-p(2 × 2) surface is then generated, as schematically shown in Figure 4d. However, as shown from its [11̅0] side view (see details in Figure S5), this surface exhibits two vacant columns for every three atomic columns, which is inconsistent with experimental results as shown in Figure 3. Therefore, the (111) rock-salt surface cations must exhibit a disordered configuration, as shown in Figure 4e. The [11̅0] side view of the disordered rock-salt surface cations (Figure 4f) has no vacant columns, while the lower concentration of the surface cations compared to the bulk leads to their reduced contrast, consistent with the experimental data in Figure 3.

Figure 5. (a) STEM view showing a (002) rock-salt reconstruction on a pristine primary particle. The TM cations migrate along the lithium channels toward the bulk of the particle, inducing the formation of the rock-salt phase. (b) Magnified view of the bulk area in (a), which is the typical layered phase with no Li-site TM cations. (c) Magnified view of the subsurface area in (a), which shows a slight contrast of Li-site TM cations. (d) Magnified view of the surface area in (a), where the Li-site and TM-site TM cations have almost the same contrast, confirming that it has a rock-salt structure.

■

little to the HADDF contrast,39,40 so the Li sites exhibit a black contrast. In the subsurface region (Figure 5c), Li sites exhibit a dim contrast due to a partial presence of TM cations on the Li sites. The contrast of the Li and TM sites becomes the same in a disordered rock-salt structure. Therefore, the subsurface area still preserves the layered structure, with rock-salt features due to the presence of TM cations on Li sites.38,41 An equal occupancy of all sites by TM cations leads to the formation of a completely disordered rock-salt phase on the particle surface, as shown in Figure 5d. The depth-dependent phase distribution shown in Figure 5b−d indicates that the rock-salt transformation is a function of the inward diffusion of TM cations. Details of the relationship between the layered → rock-salt transformation and the STEM-HAADF contrast are presented in Figure S6. The layered → rock-salt phase transformation determines the (002) reconstruction shown in Figure 5. In the NCA cathode,

DISCUSSION Reconstructing Mechanisms of the (002) and (111) Rock-Salt Surfaces. As aforementioned, two different facets of the rock-salt layer are present on the pristine primary particle: (002) and (111). Details of the atomic processes leading to the two surface configurations are presented below. Figure 5 indicates that the (002) rock-salt reconstruction forms due to gradual migration of TM cations from the surface region into the subsurface area, using the lithium ion channels (marked in Figure 5a) as the migration pathway due to the small size of Li+ and its loose attachment to the lattice.6,37,38 Magnified STEM views of the Li-channels obtained from the bulk region to the surface clearly illustrate the migration pathway of TM cations (Figure 5b−d). The bulk of the particle maintains the layered structure with no TM cations on the Lisites, as demonstrated in Figure 5b. Li+ cations contribute very 695

DOI: 10.1021/acs.chemmater.7b03901 Chem. Mater. 2018, 30, 692−699

Article

Chemistry of Materials

Figure 6. (a) HRTEM view of the atomically rough (014̅) surface of the layered phase before the (002) rock-salt reconstruction. (b) Schematic showing the atomically rough (014)̅ surface of the layered structure without the (002) rock-salt reconstruction, comparable to (a). (c) Schematic showing the interior migration of the protruding TM cations on the (014̅) surface of the layered phase leading to the (002) rock-salt reconstruction. (d) Schematic showing the reconstructed (002) rock-salt surface with random, positively charged TM cations remaining on the surface. (e) Schematic showing the unstable, exposed TM cations migrating into the subsurface region of the particle. (f, g) Schematics showing the merging of the exposed TM cations on the (002) rock-salt surface to form a stepped structure, minimizing the total free surface area of the exposed cations. The positively charged free surfaces in (f) and (g) are marked with orange dashed lines.

Figure 7. (a) STEM image showing the inward development of the (111) rock-salt reconstruction along the c-axis of the layered phase, forming a straight reconstruction surface. (b) Interlayer mixing which transforms the Li+ cations onto the (003) surface of the layered phase. (c) Schematic showing partial interlayer mixing of the Li/TM cations occurring on the (003) surface of the primary particle, leading to the (111) rock-salt reconstruction with a flat surface. (d) Complete interlayer mixing in which Li+ detaches from the (003) surface and TM cations migrate along the Li channels, leading to the (111) reconstruction. (e) Stepped surface resulting from the (111) reconstruction. (f) Schematic showing the zigzag surface resulting from the (111) rock-salt reconstruction.

the rock-salt phase has the formula of MO, in which M = ∼70− 100% Ni2+ cations,42 along with a small amount of Li, Co, and Al cations.33 In other words, the rock-salt phase is enriched with TM cations (nickel in this case) compared with the layered phase. The inward migration of TM cations increases their concentration, providing the necessary condition for the rock-salt transformation. Accompanying the enrichment of TM cations, lithium and oxygen are partially removed from the

surface area, escaping into the environment or forming lithium salts such as hydroxide or carbonate.41 Knowing from Figure 5 that the (002) reconstruction develops via the inward migration of TM cations onto the Li sites, a detailed pathway for the (002) rock-salt reconstruction is proposed, as demonstrated in Figure 6. The (002) plane of the rock-salt phase corresponds to the (014̅) plane of the layered phase, on which the (002) rocksalt reconstruction takes place. HRTEM observation of a clear 696

DOI: 10.1021/acs.chemmater.7b03901 Chem. Mater. 2018, 30, 692−699

Article

Chemistry of Materials (014̅ ) surface of the layered phase without rock-salt reconstruction shows a rough morphology (Figures 6a and S7). As schematically shown in Figure 6b, this layered surface is atomically rough with multiple protruding TM cations, where the TM cations in the surface have fewer nearest neighbors and thus have dangling bonds exposed to the surface. These surface dangling bonds produce an inwardly directed force that drives the migration of the surface atoms into Li sites in the subsurface region, as shown in Figure 6c. This inward migration leads to the enrichment of TM cations in the surface and subsurface region, transforming the rough (014̅) surface of the layered phase into the flat (002) surface of the rock-salt phase. During this process, the oxygen and lithium ions in the surface and subsurface regions are partially removed to enable the rock-salt transformation induced by the inward migration of TM cations. The high instability of the rough (014̅) surface accelerates the inward diffusion of the TM cations and, by extension, the (002) rock-salt reconstruction. After the (002) reconstruction, any individually exposed TM cations on the surface, as schematically shown in Figure 6d, render it unstable through the formation of uncompensated charge. These defects tend to be eliminated to minimize the surface energy via the formation of stepped structures. Theoretically, there are two possible reconstruction pathways to eliminate the unstable, exposed TM cations on the (002) reconstruction to minimize the surface free energy, as schematically shown in Figure 6e−g. The first one is through inward migration of exposed cations into the subsurface region of the parent particle (the interior of the (002) reconstruction), leading to the inward growth of the rock-salt layer, as schematically illustrated in Figure 6e. The other pathway involves the aggregation of the highly exposed cations on the (002) surface, as illustrated in Figures 6f,g and S8. The aggregation of the exposed cations results in the formation of atomic steps on the (002) surface of the rock-salt layer, consistent with the experimental observations. This mechanism lowers the total polar surface area of exposed cations (see orange dashed lines in Figure 6f,g) and reduces electrostatic energy. Unlike the growth of the (002) reconstruction along the lithium channels, the (111) reconstruction propagates inward along the c-axis of the layered phase, which is perpendicular to the lithium channels (Figure 7a). The topmost area is fully transformed into the rock-salt phase, the subsurface is a partially developed rock-salt phase, and the bulk preserves a rather defect-free layered framework. In the absence of lithium channels, this (111) reconstruction develops via the interlayer mixing of the lithium and TM cations, as schematically shown in Figure 7b. Li/Ni interlayer mixing is facile in the Ni-rich cathode studied here,7,9,11,16,38 which transforms the R3̅m layered structure into the Fm3̅m rock-salt structure with a much lower Gibbs free energy.38,43 The roughening effect of the (111) reconstruction depends on the development of the interlayer mixing. When a small amount of the TM cations exchanges the position with Li+ (Figure 7c), the framework of the layered phase is maintained and the (111) reconstruction exhibits a straight surface (Figure 7a). When the TM and Li cations completely shift their positions, the Li cations become the outermost surface atomic layer while the TM cations occupy the second atomic layer (Figure 7d). The Li+ cations in the surface are highly unstable due to their increased dangling bonds, more open structure, and lower bonding energy with the oxygen anions,7,9 favoring

their extraction from the primary particle. The subsurface TM cations can migrate along the lithium channels toward the regions without interlayer mixing (Figure 7d), leading to the formation of the rock-salt phase in these regions. As a result, both lithium and TM cations in the interlayer-mixing region diffuse away, leaving behind a (111) rock-salt stepped surface structure, as schematically illustrated in Figure 7e. Such an atom migration process shown in Figure 7d,e results in the hill-andvalley surface morphology of the (111) reconstruction of the rock-salt phase, as shown in Figures 3 and 7f. In other words, the macroscale roughness of the (111) reconstruction is attributed to its roughness at the atomic level (Figure 7d,e). This hill-and-valley surface configuration formed via lattice dissociation results in lattice distortion or even amorphization in some regions of the rock-salt layer, as shown in Figure 3d. Rock-Salt Reconstruction and Electrochemistry. As shown below, the surface layered → rock-salt phase transformation depends on the structural anisotropy of the layered phase, for which the different facets of the primary particles lead to significantly different features of the surface rock-salt layer, including the growth pathways and the atomic configurations. A shared feature of the (002) and (111) rock-salt reconstructions is the loss of lithium and oxygen, leading to the enrichment of TM cations in the surface region. The ionic radius of Ni2+ cations makes them energetically favored to occupy the Li octahedral sites, leading to the formation of the rock-salt phase. In contrast, Ni3+ is much less likely to appear in interlayer defects7 because of the size mismatch with Li+. Therefore, the loss of O2− from the structure is accompanied by the reduction of the Ni3+ cations and the formation of Ni2+, which is the essential component for interlayer mixing and thus formation of the rock-salt phase. Surface rock-salt reconstructions are electrochemically inactive because they reduce the electronic and, especially, ionic conductivity of the cathode through the elimination of percolation channels for charge carriers compared to the layered phase.37 Unfortunately, the high chemical activity of the layered phase makes it very facile to transform to the rock-salt phase during heating or in contact with the gaseous environment.7,9 The limited thickness of both surface reconstructions in pristine materials (∼5 atomic layers) seems to be insufficient to shut down transport, as this oxide is highly electrochemically active.44 However, when the cycle number increases, the rock-salt reconstruction layer is found to grow into larger electrochemically inactive domains, which eventually degrades performance.10,45 In terms of atomic configuration, the (111) reconstruction is much rougher compared with the (002) reconstruction. The (002) reconstruction generates a flat, well-structured facet, while the (111) reconstruction forms an etched, loosely reconstructed surface. In the corroded areas (valley-like structures) of the (111) reconstruction, the topmost surface of the layered phase is completely etched out, and the layered structure beneath is exposed to the degrading environment. This newly exposed layered phase transforms to the rock-salt phase via the (111) reconstruction, which leads to more etching and the formation of more exposed layered phase. The chain reaction continues until the entire (003) surface of the layered phase is transformed to the (111) reconstruction, which exhibits a zigzag surface morphology resulting from the surface etching. On the contrary, the (002) reconstruction results from inward diffusion of TM cations, followed by the stabilization of the surface reconstruction. Therefore, the (111) reconstruction 697

DOI: 10.1021/acs.chemmater.7b03901 Chem. Mater. 2018, 30, 692−699

Article

Chemistry of Materials ORCID

results in a continuous formation of the rock-salt phase with a defective lattice, while the (002) reconstruction leads to the formation of a well-structured rock-salt layer with a relatively intact lattice that has improved passivation property and provides protection to the underlying layered phase. The zigzag (111) reconstruction surface significantly increases the surface area, which can lead to more cation dissolution compared with the intact, flat (002) rock-salt surface during electrochemical cycling. In other words, the different configurations of (014)̅ and (003) planes lead to (002) and (111) rock-salt reconstructions which should show significantly different chemical reactivity. From the perspective of electrochemical properties, the (002) rock-salt reconstruction blocks the lithium channels, which reduces the Li+ and electronic conductivity of the particle, an undesirable outcome. However, as mentioned above, this (002) reconstruction needs the Li+ cations to be removed for the inward diffusion of TM cations. Once the (002) reconstruction forms, the loss of Li cations becomes difficult, prohibiting the further development of the reconstruction. Conversely, the (111) reconstruction does not require the removal of Li+ cations first. The interlayer mixing of lithium and TM cations exposes the lithium cations onto the surface and leads to its detachment. As the (111) reconstruction takes place, fresh layered phase is continuously exposed in the surface (Figure 7e), which maintains the progression of (111) reconstruction. Therefore, the (111) reconstruction develops faster than the (002) reconstruction, resulting in larger electrochemically inactive rock-salt domains that reduce the ionic and electronic conductivity of the cathode.

Hanlei Zhang: 0000-0001-6540-0556 Montse Casas-Cabanas: 0000-0002-9298-2333 Jordi Cabana: 0000-0002-2353-5986 Chongmin Wang: 0000-0003-3327-0958 Guangwen Zhou: 0000-0002-9243-293X Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported as part of the NorthEast Center for Chemical Energy Storage (NECCES), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, under Award No. DESC0012583. This research used resources of the Advanced Photon Source, a U.S. Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory, under Contract No. DEAC02-06CH11357. C.W. thanks the support of the Assistant Secretary for Energy Efficiency and Renewable Energy, Office of Vehicle Technologies of the U.S. Department of Energy, under Contract No. DE-AC02-05CH11231, Subcontract No. 18769 and No. 6951379 under the Advanced Battery Materials Research (BMR) program. The TEM work was conducted in the William R. Wiley Environmental Molecular Sciences Laboratory (EMSL), a national scientific user facility sponsored by DOE’s Office of Biological and Environmental Research and located at PNNL. PNNL is operated by Battelle for the Department of Energy under Contract DE-AC05-76RLO1830.

■

■

CONCLUSION Two kinds of rock-salt induced surface reconstructions, (002)type and (111)-type, are observed on Ni-rich layered primary particles. Both reconstructions feature stepped structures, but the (002) stepping is significantly flatter compared with the (111) surface that shows a highly rough hill-and-valley morphology. The (002) reconstruction is induced by inward migration of surface TM cation along lithium channels, while the (111) reconstruction occurs due to the interlayer mixing of Li/TM cations. Both reconstructions are potential sources for structural degradation during electrochemical cycling: the (002) reconstruction reduces the ionic and electronic conductivity. In turn, the (111) reconstruction leads to significant roughing of the (003) surface which largely increases the surface area and may induce more surface etching and cation dissolution by the electrolyte.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.chemmater.7b03901. Chemistry of the pristine sample; repeated STEM/ HRTEM observations of the surface and bulk; STEM simulations; diffractograms; and schematics showing the atomic processes (PDF)

■

REFERENCES

(1) Whittingham, M. S. Ultimate Limits to Intercalation Reactions for Lithium Batteries. Chem. Rev. 2014, 114, 11414−11443. (2) Hausbrand, R.; Cherkashinin, G.; Ehrenberg, H.; Gröting, M.; Albe, K.; Hess, C.; Jaegermann, W. Fundamental Degradation Mechanisms of Layered Oxide Li-ion Battery Cathode Materials: Methodology, Insights and Novel Approaches. Mater. Sci. Eng., B 2015, 192, 3−25. (3) Whittingham, M. S. Lithium Batteries and Cathode Materials. Chem. Rev. 2004, 104, 4271−4302. (4) Yoshio, M.; Brodd, R. J.; Kozawa, A. Lithium-Ion Batteries; Springer: 2009; Vol. 1. (5) Arico, A. S.; Bruce, P.; Scrosati, B.; Tarascon, J.-M.; Van Schalkwijk, W. Nanostructured Materials for Advanced Energy Conversion and Storage Devices. Nat. Mater. 2005, 4, 366−377. (6) He, P.; Yu, H.; Zhou, H.; Li, D. Layered Lithium Transition Metal Oxide Cathodes Towards High Energy Lithium-Ion Batteries. J. Mater. Chem. 2012, 22, 3680−3695. (7) Yuan, X.; Liu, H.; Zhang, J. Lithium-Ion Batteries: Advanced Materials and Technologies; CRC Press: 2011. (8) Hy, S.; Liu, H.; Qian, D.; Zhang, M.; Hwang, B. J.; Meng, Y. S. Performance and Design Considerations For The lithium Excess Layered Oxide Positive Electrode Materials For Lithium Ion Batteries. Energy Environ. Sci. 2016, 9, 1931−1954. (9) Myung, S.-T.; Maglia, F.; Park, K.-J.; Yoon, C. S.; Lamp, P.; Kim, S.-J.; Sun, Y.-K. Nickel-Rich Layered Cathode Materials for Automotive Lithium-Ion Batteries: Achievements and Perspectives. ACS Energy Lett. 2017, 2, 196−223. (10) Lin, F.; Markus, I. M.; Nordlund, D.; Weng, T.-C.; Asta, M. D.; Xin, H. L.; Doeff, M. M. Surface Reconstruction and Chemical Evolution of Stoichiometric Layered Cathode Materials for Lithiumion Batteries. Nat. Commun. 2014, 5, 3529. (11) Yan, P.; Zheng, J.; Zheng, J.; Wang, Z.; Teng, G.; Kuppan, S.; Xiao, J.; Chen, G.; Pan, F.; Zhang, J. G.; Wang, C.-M. Ni and Co Segregations on Selective Surface Facets and Rational Design of

AUTHOR INFORMATION

Corresponding Authors

*(C.W.) E-mail: [email protected]. *(G.Z.) E-mail: [email protected]. 698

DOI: 10.1021/acs.chemmater.7b03901 Chem. Mater. 2018, 30, 692−699

Article

Chemistry of Materials Layered Lithium Transition-Metal Oxide Cathodes. Adv. Energy Mater. 2016, 6, 1502455. (12) Lin, F.; Markus, I. M.; Doeff, M. M.; Xin, H. L. Chemical and Structural Stability of Lithium-Ion Battery Electrode Materials under Electron Beam. Sci. Rep. 2015, 4, 5694. (13) Hwang, S.; Chang, W.; Kim, S. M.; Su, D.; Kim, D. H.; Lee, J. Y.; Chung, K. Y.; Stach, E. A. Investigation of Changes in the Surface Structure of LixNi0.8Co0.15Al0.05O2 Cathode Materials Induced by the Initial Charge. Chem. Mater. 2014, 26, 1084−1092. (14) Mohanty, D.; Dahlberg, K.; King, D. M.; David, L. A.; Sefat, A. S.; Wood, D. L.; Daniel, C.; Dhar, S.; Mahajan, V.; Lee, M.; Albano, F. Modification of Ni-Rich FCG NMC and NCA Cathodes by Atomic Layer Deposition: Preventing Surface Phase Transitions for HighVoltage Lithium-Ion Batteries. Sci. Rep. 2016, 6, 26532. (15) Jung, S. K.; Gwon, H.; Hong, J.; Park, K. Y.; Seo, D. H.; Kim, H.; Hyun, J.; Yang, W.; Kang, K. Understanding the Degradation Mechanisms of LiNi0.5Co0.2Mn0.3O2 Cathode Material in Lithium Ion Batteries. Adv. Energy Mater. 2014, 4, 1300787. (16) Zhang, H.; Karki, K.; Huang, Y.; Whittingham, M. S.; Stach, E. A.; Zhou, G. Atomic Insight into the Layered/Spinel Phase Transformation in Charged LiNi0.80Co0.15Al0.05O2 Cathode Particles. J. Phys. Chem. C 2017, 121, 1421−1430. (17) Lin, F.; Nordlund, D.; Li, Y.; Quan, M. K.; Cheng, L.; Weng, T.C.; Liu, Y.; Xin, H. L.; Doeff, M. M. Metal Segregation in Hierarchically Structured Cathode Materials for High-Energy Lithium Batteries. Nat. Energy 2016, 1, 15004. (18) Wei, Z.; Zhang, W.; Wang, F.; Zhang, Q.; Qiu, B.; Han, S.; Xia, Y.; Zhu, Y.; Liu, Z. Eliminating Voltage Decay of Lithium-Rich Li1.14Mn0.54Ni0.14Co0.14O2 Cathodes by Controlling the Electrochemical Process. Chem. - Eur. J. 2015, 21, 7503−7510. (19) Fell, C. R.; Carroll, K. J.; Chi, M.; Meng, Y. S. Synthesis− Structure−Property Relations in Layered,“Li-excess” Oxides Electrode Materials Li[Li1/3−2x/3NixMn2/3−x/3]O2 (x= 1/3, 1/4, and 1/5). J. Electrochem. Soc. 2010, 157, A1202−A1211. (20) Song, B.; Liu, Z.; Lai, M. O.; Lu, L. Structural Evolution and the Capacity Fade Mechanism upon Long-Term Cycling in Li-Rich Cathode Material. Phys. Chem. Chem. Phys. 2012, 14, 12875−12883. (21) Watanabe, S.; Kinoshita, M.; Hosokawa, T.; Morigaki, K.; Nakura, K. Capacity Fade of LiAlyNi1−x−yCoxO2 Cathode for LithiumIon Batteries during Accelerated Calendar and Cycle Life Tests (Surface Analysis of LiAlyNi1−x−yCoxO2 Cathode after Cycle Tests in Restricted Depth of Discharge Ranges). J. Power Sources 2014, 258, 210−217. (22) Lee, S.; Yoon, G.; Jeong, M.; Lee, M. J.; Kang, K.; Cho, J. Hierarchical Surface Atomic Structure of a Manganese-Based Spinel Cathode for Lithium-Ion Batteries. Angew. Chem., Int. Ed. 2015, 54, 1153−1158. (23) Zhao, X.; Zhou, F.; Dahn, J. Phases Formed in Al-Doped Ni1/3Mn1/3Co1/3(OH)2 Prepared by Coprecipitation: Formation of Layered Double Hydroxide. J. Electrochem. Soc. 2008, 155, A642− A647. (24) Karki, K.; Huang, Y.; Hwang, S.; Gamalski, A. D.; Whittingham, M. S.; Zhou, G.; Stach, E. A. Tuning the Activity of Oxygen in LiNi0.8Co0.15Al0.05O2 Battery Electrodes. ACS Appl. Mater. Interfaces 2016, 8, 27762−27771. (25) Robert, R.; Bünzli, C.; Berg, E. J.; Novák, P. Activation Mechanism of LiNi0.80Co0.15Al0.05O2: Surface and Bulk Operando Electrochemical, Differential Electrochemical Mass Spectrometry, and X-ray Diffraction Analyses. Chem. Mater. 2015, 27, 526−536. (26) Liu, B.-s.; Wang, Z.-B.; Yu, F.-D.; Xue, Y.; Wang, G.-j.; Zhang, Y.; Zhou, Y.-X. Facile Strategy of NCA Cation Mixing Regulation and its Effect on Electrochemical Performance. RSC Adv. 2016, 6, 108558−108565. (27) Tamura, A.; Takai, S.; Yabutsuka, T.; Yao, T. Relaxation Analysis of LixNiO2 and Lix(NCA)O2 in the Deeply Lithium Extracted Region (x ≤ 0.12). J. Electrochem. Soc. 2017, 164, A1514−A1519. (28) Boulineau, A.; Simonin, L.; Colin, J.-F.; Bourbon, C.; Patoux, S. First Evidence of Manganese−Nickel Segregation and Densification

upon Cycling in Li-Rich Layered Oxides for Lithium Batteries. Nano Lett. 2013, 13, 3857−3863. (29) Hannay, N. B. Treatise on Solid State Chemistry; Plenum Publishing Corporation: 1973; Vol. 6. (30) Chan, K. S.; Munson, D. E.; Bodner, S. R.; Fossum, A. F. Cleavage and Creep Fracture of Rock Salt. Acta Mater. 1996, 44, 3553−3565. (31) Bassett, G. A New Technique for Decoration of Cleavage and Slip Steps on Ionic Crystal Surfaces. Philos. Mag. 1958, 3, 1042−1045. (32) Zhang, H.; Omenya, F.; Whittingham, M. S.; Wang, C.; Zhou, G. Formation of an Anti-Core−Shell Structure in Layered Oxide Cathodes for Li-Ion Batteries. ACS Energy Lett. 2017, 2, 2598−2606. (33) Goodenough, J.; Wickham, D.; Croft, W. Some Magnetic and Crystallographic Properties of the System Li+xNi++1−2xNi+++xO. J. Phys. Chem. Solids 1958, 5, 107−116. (34) Watanabe, S.; Kinoshita, M.; Nakura, K. Capacity Fade of LiNi(1−x−y)CoxAlyO2 Cathode for Lithium-Ion Batteries during Accelerated Calendar and Cycle Life Test. I. Comparison Analysis between LiNi(1−x−y)CoxAlyO2 and LiCoO2 Cathodes in Cylindrical Lithium-Ion Cells during Long Term Storage Test. J. Power Sources 2014, 247, 412−422. (35) Ino, S. Epitaxial Growth of Metals on Rocksalt Faces Cleaved in Vacuum. II. Orientation and Structure of Gold Particles Formed in Ultrahigh Vacuum. J. Phys. Soc. Jpn. 1966, 21, 346−362. (36) Barbier, A.; Mocuta, C.; Kuhlenbeck, H.; Peters, K.; Richter, B.; Renaud, G. Atomic Structure of the Polar NiO (111)-p (2× 2) Surface. Phys. Rev. Lett. 2000, 84, 2897. (37) Lee, J.; Urban, A.; Li, X.; Su, D.; Hautier, G.; Ceder, G. Unlocking the Potential of Cation-Disordered Oxides for Rechargeable Lithium Batteries. Science 2014, 343, 519−522. (38) Yan, P.; Zheng, J.; Lv, D.; Wei, Y.; Zheng, J.; Wang, Z.; Kuppan, S.; Yu, J.; Luo, L.; Edwards, D.; et al. Atomic-Resolution Visualization of Distinctive Chemical Mixing Behavior of Ni, Co and Mn with Li in Layered Lithium Transition Metal Oxide Cathode Materials. Chem. Mater. 2015, 27, 5393−5401. (39) Findlay, S. D.; Huang, R.; Ishikawa, R.; Shibata, N.; Ikuhara, Y. Direct Visualization of Lithium via Annular Bright Field Scanning Transmission Electron Microscopy: A Review. Microscopy 2017, 66, 3−14. (40) Williams, D. B.; Carter, C. B. Transmission Electron Microscopy: A Textbook for Materials Science; Springer: 2009. (41) Sallis, S.; Pereira, N.; Mukherjee, P.; Quackenbush, N.; Faenza, N.; Schlueter, C.; Lee, T.-L.; Yang, W.; Cosandey, F.; Amatucci, G.; Piper, L. F. J. Surface Degradation of Li1−xNi0.80Co0.15Al0.05O2 Cathodes: Correlating Charge Transfer Impedance with Surface Phase Transformations. Appl. Phys. Lett. 2016, 108, 263902. (42) Sasaki, S.; Fujino, K.; Takéuchi, Y. X-ray Determination of Electron-Density Distributions in Oxides, MgO, MnO, CoO, and NiO, and Atomic Scattering Factors of Their Constituent Atoms. Proc. Jpn. Acad., Ser. B 1979, 55, 43−48. (43) Conry, T. E.; Mehta, A.; Cabana, J.; Doeff, M. M. Structural Underpinnings of the Enhanced Cycling Stability upon Al-Substitution in LiNi0.45Mn0.45Co0.1−yAlyO2 Positive Electrode Materials for Li-ion Batteries. Chem. Mater. 2012, 24, 3307−3317. (44) Chen, C.; Liu, J.; Stoll, M.; Henriksen, G.; Vissers, D.; Amine, K. Aluminum-Doped Lithium Nickel Cobalt Oxide Electrodes for HighPower Lithium-Ion Batteries. J. Power Sources 2004, 128, 278−285. (45) Yang, F.; Liu, Y.; Martha, S. K.; Wu, Z.; Andrews, J. C.; Ice, G. E.; Pianetta, P.; Nanda, J. Nanoscale Morphological and Chemical Changes of High Voltage Lithium−Manganese Rich NMC Composite Cathodes with Cycling. Nano Lett. 2014, 14, 4334−4341.

699

DOI: 10.1021/acs.chemmater.7b03901 Chem. Mater. 2018, 30, 692−699