Tunable Single-Atomic Charge/Spin States of Intercalant Co Ions in a Transition Metal Dichalcogenide

doi:10.21203/rs.3.rs-647929/v1

Download PDF

Article

Tunable Single-Atomic Charge/Spin States of Intercalant Co Ions in a Transition Metal Dichalcogenide

https://doi.org/10.21203/rs.3.rs-647929/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Intercalation raises manifold possibilities to manipulate the properties of two-dimensional (2D) materials¹, and its impact on local electronic/magnetic properties has drawn much attention with the rise of nano-structured 2D materials^2,3. Typically, changing an ionic state in a solid involves a dramatic local change of energy as well as orbital/spin magnetic moment from its ground state. However, the atomic investigation of the charging process of an intercalant ion in 2D material has never been explored while such subject has been studied in artificially deposited atoms on thin insulating 2D layers using scanning probe microscopy^4–7. Herein, we demonstrate an atomical manipulation of the charge and spin state of Co ions on a metallic NbS₂, obtained by cleaving of Co-intercalated NbS₂. Density functional theory investigation of various Co configurations reveals that the charging is possible due to a change in the crystal field at the surface and a significant coupling between NbS₂ and intercalants occurs via orbitals of the a_1g symmetry. The results can be generalized to numerous other combinations of intercalants and base matrixes, suggesting that intercalated transition metal dichalcogenides can be a new platform to introduce single-atom operation 2D electronics/spintronics.

Hard Condensed-matter Physics

Soft Condensed-matter Physics

Materials Theory and Modeling

two-dimensional (2D) materials

intercalation

electronics

spintronics

Intercalation has long been studied as a strategy to derive various properties from graphene⁸, and the examples span from the practical use of lithium intercalation in battery⁹ to the non-trivial intercalation-induced superconductivity^10,11. With the advent of transition metal dichalcogenides (TMD) in 2D spintronics¹² as a bridging material between two principal components, i.e. graphene and hexagonal boron nitride, intercalation has been adopted in TMD as well to derive various emergent properties, such as generation of magnetic order with self-intercalation¹³, controlled doping³, rectification of thermal conduction¹⁴, catalytic efficiency in hydrogen evolution reaction¹⁵, Z₃ Potts-nematic antiferromagnetism¹⁶, and electrically switchable magnetic order¹⁷. Despite the diverse combinations of intercalant and base matrix, TMDs intercalated with 3d transition metals are of particular interest due to their great tunability in various magnetic ground states such as ferromagnetic^18–20, antiferromagnetic²¹, or helical magnetism^22,23. They have several distinguished properties such as the localized intercalant d-orbital state within the delocalized state of the matrix^24,25 and distinguished magnetic critical behavior²⁶. Moreover, remarkable properties such as anomalous hall effect^20,21,27 or spontaneous spin-polarized electric conduction²⁸ suggest potential discovery of non-trivial phenomena with atomic investigation of intercalants in 3d metal intercalated TMDs.

In this report, we show that a Co ion on a cleaved surface of intercalated Co_1/3NbS₂ can exhibit a metastable charged state distinguished from other structural defect states. The metastable charged state can be controlled using scanning tunneling microscopy (STM) probe, which resulted in manipulation of charge and spin states of Co ion. Although the process is similar to what have been reported in single electron charging of deposited atom on thin insulating film⁴, the absence of insulating layer_, which is the essential part of various charge manipulation experiments^4–7, contrasts the fundamental difference of the process. The existence of van der Waals gap in the bulk structure seems involved in reducing coupling of the intercalated ion orbitals from the base matrix states even without any further insulating layer in between. We have performed density functional theory (DFT) calculations to understand the intriguing existence of the metastable charged state on the metallic substrate and revealed that the reduction of crystal field on the surface and subsequent shift of localized a_1gorbital towards the Fermi level play a role in making the localized state accessible by the STM tip. The result demonstrates the smallest structure in which a single charged ion can exist on a metallic compound by adopting intercalation into 2D TMD material and can be a new pathway to incorporate electronic/magnetic control over various 2D materials.

Co_1/3NbS₂ is an intercalated version of metallic NbS₂ where the Co ions occupy one third of Nb atop sites within the van der Waals gap as in Fig. 1a. The intercalant position alternates between two columns along c axis which produces two thirds of Nb ions in Wyckoff 4f sites while one third of Nb sites remain 2a sites. As a result, Co ions form a triangular lattice within one van der Waals gap and they tend to remain on one surface when the crystal is cleaved at low temperature (see methods for details). For convenience, we will refer each layer from the cleaved surface towards bulk with numbers as indicated in Fig. 1a, e.g. layer 1 Co, layer 2 S, layer 3 Nb, etc. Fig. 1b shows an STM topography of a cleaved Co_1/3NbS₂ surface terminated with Co. In comparison with the simulated STM image showing the ideal trigonal lattice of Co ions (Fig. 1b inset), the STM image shows the existence of various types of defects. Exhaustive DFT calculations make it possible to determine the exact atomic structure of all defect types. The most frequent defects are single/double Co vacancies in layer 1 (Fig. 1c and d), Co substitution of Nb in layer 3 (Fig. 1e), and Co trivacancy in layer 5 (Fig. 1f). Simulated STM images for described defects in Fig. 1c-f as well as other structural defect types are presented in Supplementary Information (SI, discussion in SI section 2).

We should pay additional attention to one feature in STM images, shown in Fig. 1g, which deserves a special consideration. First, it does not fit to any structural model and more importantly it is metastable and degrades with time (see below). It is clearly distinguished from other defects in the extent of vertical/lateral expansion as indicated in Fig. 1h (red solid circle: metastable state, Fig. S13 in SI). We explored the origin of these metastable states by means of DFT+U+(SOC) calculations, which also provide other important information. E.g. they show that the electronic structure of Co in layers 1 (i.e. at surface) and 5 (internal layer) is very different (Fig. 2a). The latter has 2+ (3d⁷) high-spin state with S=3/2 (Fig. 2d) similar to the reported Co valency of 2+ in the bulk measurement²⁹. It is in contrast to what we have at the surface. Detailed analysis of occupation matrix unambiguously demonstrates that Co in layer 1 is 1+ (3d⁸, S=1) with additional electron at the a_1g orbital (Fig. 2e) and its local magnetic moment is 1.93 . This is due to the specific surface termination where only three of S ions remain out of initial CoS₆ octahedron. As a result the t_2g-e_g crystal-field splitting for surface Co practically vanishes: projecting DFT Hamiltonian onto the Co 3d Wannier-functions³⁰ we find that it reduces from ~1.2 eV in bulk down to ~0.3 eV for Co at the surface. The reduction stimulates occupation of the Co a_1g states and leads to redistribution of electrons at the surface, so that these Co turn out to be closer to 1+ oxidation state. Importantly the hybridization between Nb 3z²-r² and Co a_1g orbitals (directed to each other as in Fig. 2b) pins the latter at the Fermi level and makes it accessible for STM.

Moreover, while surface Co ions are 1+ in the ground state, the Co²⁺high-spin configuration is also locally stable, and it is rather close in energy. The total energy of such metastable state is only 130 meV/Co higher than the ground state. Other possible metastable configurations are low-spin (S=0) Co¹⁺and (S=1/2) Co²⁺states; however, analysis of calculations shows that such states are locally unstable (stable only at corresponding constraints) and energetically much higher than the ground state. The instability of these configurations can be naturally explained by a large intra-atomic Hund’s exchange for Co (J_H~1 eV³¹) and non-zero exchange field due to Co ions in deeper layers.

It is important that the metastable state of surface Co with the different charge and magnetic state is not only locally stable, but also in the energy range accessible for its switching by STM. To check this hypothesis, we used a STM tip with a combination of various tip heights and biases to induce the metastable surface Co²⁺ state. Fig. 3a illustrates the optimal charging procedure that could successfully induce the metastable charged state. After locating the tip on top of the target Co ion, the tip height is lowered by changing setpoint current to 100 pA which is higher than the normal imaging condition, 10 pA at 5 mV. It is followed by a bias voltage pulse to the tip which is kept at 150 mV for 300 ms indicated by the red stars in Fig. 3a.

We could confirm creation of the metastable charged state by subsequent tunneling current change as well as from STM images showing created Co²⁺ ions having the same height profile with the naturally found ones. Fig. 3b-g show a series of STM images taken on an area while we apply pulses in between. The points marked with red stars in each image were applied a voltage pulse immediately right after, and then the targeted Co ion becomes a metastable charged Co²⁺ showing bright protrusion in the next image. For example, we applied the pulse to the red starred point in Fig. 3b right after the image, then the following Fig. 3c shows a bright metastable charged state at the position. Similarly, we used two pulses after Fig. 3c that resulted in producing the two bright Co ions in Fig. 3d, and so on. One interesting point is that the first charged Co has returned to the ground state between Fig. 3e and f (black dashed circle), but the point was charged again by applying another pulse after Fig. 3f, which demonstrates its bistable nature.

To estimate the overall lifetime of the charged state, we have performed a charging experiment on a massive number of Co sites, and it was estimated to be 43 min (SI section 3.1). The measured lifetime is very unusual compared to Au charging on NaCl/Cu³² that requires two or more insulating layers to obtain an hour-long lifetime, or donor state in a semiconductor that exists only under tip induced band (i.e., zero lifetime)³³. The discovery of the intriguing role of van der Waals gap in inducing orbital separation between intercalant and metallic base can be of particular importance in this study. Besides that, we were able to see that the lifetime of the charged state varies a lot by its environment, and the extreme case is when the charged state is in the vicinity of Co vacancies where they show frequent changes within a single frame of STM image as shown in red dashed circles in Fig. 1h. It implies that an even longer-lived charged state can be obtained by controlling the number of structural defects in the crystal. The result opens the possibility of switching the Co d⁷-d⁸ states³⁴ with unprecedented lifetime on a conducting substrate.

Another unique aspect of the charging process is that we could induce the change in both of bias polarities as shown schematically in Fig. 4a. When the sample bias is positive, a single electron hopping happens from the Co ion to the sample as the direction of tunneling electrons. On the other hand, when the sample bias is negative, the direction of tunneling current is reversed and likewise the single electron hopping occurs towards the tip. However, the change of Co is always from Co¹⁺ to Co²⁺ by losing one electron. It is in contrast with the case of single atom on insulating NaCl layer^4,35 which shows suppressed electron hopping towards the sample by the presence of insulating layer and gets an electron only from tip. Thus, the bi-polarity nature of the process is another illustration of the fact that the localized Co 3d electrons at surface are coupled somewhat with the rest of metallic sample. In order to compare the electron hopping probability towards tip and sample, we compared the time to generate a charged Co²⁺ in both polarities. (SI Table S1) The estimated numbers of tunneling electrons to provoke charged state are measured to be and electrons for the tunneling towards tip and sample, respectively (at normalized tunneling current of 300 pA with sample biases of ±200 mV). Considering that the hopping rate towards the tip can be tuned by adjusting the vacuum gap, those comparable rates imply the possibility of fine tuning between electron hopping towards tip and sample.

Lastly, we demonstrate formation of a linear chain of S=3/2 Co²⁺ ions through Fig. 4b-g. Fig. 4g shows the final result of eight consecutive S=3/2 Co²⁺ chain and interestingly the charging pulses were applied twice for the first and seventh Co ions as they have changed back to the ground state during the process. In principle, the procedure can be adopted to arrange S=3/2 Co²⁺ ions of any desired shape in background S=1/2 Co¹⁺ lattice unless the system has low enough defect density. As the procedure can arrange ions with different magnetic moment freely, a cleaved surface of Co_1/3NbS₂ can be an interesting playground to study the interactions among various arrangements of ions with different magnetic moments.

In summary, we have shown that one can atomically tune the charge and spin state of intercalant Co ions from a cleaved surface of Co_1/3NbS₂ by STM. It is possible because of the reduced t_2g-e_g crystal-field splitting of surface Co as well as the decoupled nature of the intercalant orbitals from its matrix. The metastable Co²⁺ high-spin state turns out to be long-lived and transforms to ground state up to a time scale of ~ 43 minutes, which is extraordinarily long for an ion placed on a metallic substrate. In contrast to the previously reported single atom-type charging processes a whole lattice of switchable Co ions is available in the case of Co_1/3NbS₂ and its combination with metallic internal layers can be potentially useful for spintronics. Finally, Co_1/3NbS₂ is just one example of a large family of intercalated TMDs, which can have a different intercalated transition metal, varying degree of intercalation, or a choice of intercalation host, with a potentially similar atomic tunability with different combinations of spin and charge states.

Crystal growth

Single crystalline Co_1/3NbS₂ samples were synthesized by chemical vapor transport method using iodine transport agent. Co rich molar ratio of Co:Nb:S=7:10:20 was used to tune the final composition to Co_1/3NbS₂ characterized by energy dispersive x-ray analysis. The 0.2 g of mixture together with 0.1 g of iodine were sealed in a quartz tube under vacuum and placed in a two-zone furnace. The set point temperatures of the two zones were 800^⁰C and 950^⁰C, which were kept for 4 weeks. Hexagonal shaped single crystals were collected after cooling the furnace to room temperature.

Scanning Tunneling Microscopy

Co_1/3NbS₂ samples were fixed to a CuBe sample plate and a metal post was attached to the top surface with Epotek H20E epoxy. The cured samples were transferred to the ultra-high vacuum chamber and inserted to the cleaving stage. The sample was cooled at the stage by liquid nitrogen for 30 mins prior to cleaving and moved to the STM head right after removal of the metal post. STM measurements were performed using USM-1500 SPM system from Unisoku operated with Nanonis controller. The tips were characterized on sputter/anneal cleaned Cu(111) sample prior to use. All the STM measurements are measured at 78 K.

DFT Calculation and STM Simulation

The investigation of the Co_1/3NbS₂ was carried out within the framework of density functional theory (DFT)³⁶ implemented in computational package VASP^37–39. The generalized gradient approximation in form of Ref. 40 was used. All calculations were carried out with the use of the projector-augmented wave (PAW)⁴¹ basis set technique. To consider strong electronic correlations GGA+U⁴² method was used. We chose the on-site Coulomb repulsion parameter for Co to be U = 6 eV, while the Hund’s rule coupling parameter (J_H) was taken as J_H = 1 eV³¹. The energy cutoff of plane waves was chosen to be 500 eV. The integration over the first Brillouin zone was performed according to Monkhorst-Pack scheme⁴³ with the density of k-point being equal to 0.014 2π/Å. The electron population numbers were obtained by integration within atomic spheres with radii 1.302 Å, 1.503 Å, and 1.164 Å for Co, Nb and S correspondingly around each atom. The atomic structure relaxation was carried out until the maximum value of interatomic forces became less than 5∙10^-3 eV/Å, and the energy variation between two steps of electronic loop became less than 1∙10^-5 eV. To avoid an artificial interaction between structures in non-periodic directions a vacuum region of at least 15 Å was chosen. The simulated STM images of all considered structures were obtained by Hive software⁴⁴ based on the Tersoff-Hamann approximation⁴⁵. For more calculation details see Supplementary Information Section 1.

Author contributions

S.L. performed the STM measurement. S.P. and K.W. were responsible for growth. A.V.U., S.V.S, and E.V.S. performed DFT calculations, E.V.S., Z.I.P., and D.G.K. carried out the STM simulations. S.L., S.V.S., and S.W.C. co-wrote the paper, and all authors revised the manuscript. All authors discussed the results and S.W.C. supervised the whole work.

Acknowledgement

This work was supported by the center for Quantum Materials Synthesis (cQMS), funded by the Gordon and Betty Moore Foundation’s EPiQS initiative through grant GBMF6402 and by Rutgers University. DFT calculations were supported by the Russian Science Foundation via RSF 20-62-46047 project, while simulation of STM patterns was funded by RFBR and SC RA, project number 20-53-05009.

1. Wang, Z., Li, R., Su, C. & Loh, K. P. Intercalated phases of transition metal dichalcogenides. SmartMat 1, (2020).

2. Wan, J. et al. Tuning two-dimensional nanomaterials by intercalation: materials, properties and applications. Chem. Soc. Rev. 45, 6742–6765 (2016).

3. Gong, Y. et al. Spatially controlled doping of two-dimensional SnS2 through intercalation for electronics. Nature Nanotechnology 13, 294–299 (2018).

4. Repp, J., Meyer, G., Olsson, F. E. & Persson, M. Controlling the Charge State of Individual Gold Adatoms. Science 305, 493–495 (2004).

5. Patera, L. L., Queck, F., Scheuerer, P. & Repp, J. Mapping orbital changes upon electron transfer with tunnelling microscopy on insulators. Nature 566, 245–248 (2019).

6. Scheuerer, P. et al. Charge-Induced Structural Changes in a Single Molecule Investigated by Atomic Force Microscopy. Phys. Rev. Lett. 123, 066001 (2019).

7. Scheuerer, P., Patera, L. L. & Repp, J. Manipulating and Probing the Distribution of Excess Electrons in an Electrically Isolated Self-Assembled Molecular Structure. Nano Lett. 20, 1839–1845 (2020).

8. Dresselhaus, M. S. & Dresselhaus, G. Intercalation compounds of graphite. Advances in Physics 51, 1–186 (2002).

9. Li, Y., Lu, Y., Adelhelm, P., Titirici, M.-M. & Hu, Y.-S. Intercalation chemistry of graphite: alkali metal ions and beyond. Chem. Soc. Rev. 48, 4655–4687 (2019).

10. Csányi, G., Littlewood, P. B., Nevidomskyy, A. H., Pickard, C. J. & Simons, B. D. The role of the interlayer state in the electronic structure of superconducting graphite intercalated compounds. Nature Physics 1, 42–45 (2005).

11. Hannay, N. B. et al. Superconductivity in Graphitic Compounds. Phys. Rev. Lett. 14, 225–226 (1965).

12. Avsar, A. et al. Colloquium: Spintronics in graphene and other two-dimensional materials. Rev. Mod. Phys. 92, 021003 (2020).

13. Zhao, X. et al. Engineering covalently bonded 2D layered materials by self-intercalation. Nature 581, 171–177 (2020).

14. Li, B. et al. Liquid-like thermal conduction in intercalated layered crystalline solids. Nature Materials 17, 226–230 (2018).

15. Yang, J. et al. Ultrahigh-current-density niobium disulfide catalysts for hydrogen evolution. Nature Materials 18, 1309–1314 (2019).

16. Little, A. et al. Three-state nematicity in the triangular lattice antiferromagnet Fe 1/3 NbS 2. Nature Materials 19, 1062–1067 (2020).

17. Nair, N. L. et al. Electrical switching in a magnetically intercalated transition metal dichalcogenide. Nature Materials 19, 153–157 (2020).

18. Ko, K.-T. et al. RKKY Ferromagnetism with Ising-Like Spin States in Intercalated Fe1/4TaS2. Physical Review Letters 107, (2011).

19. Fan, S. et al. Electronic chirality in the metallic ferromagnet Fe1/3TaS2. Phys. Rev. B 96, 205119 (2017).

20. Mangelsen, S. et al. Large Anomalous Hall Effect and Slow Relaxation of the Magnetization in Fe1/3TaS2. J. Phys. Chem. C 124, 24984–24994 (2020).

21. Ghimire, N. J. et al. Large anomalous Hall effect in the chiral-lattice antiferromagnet CoNb3S6. Nature Communications 9, 3280 (2018).

22. Togawa, Y. et al. Chiral Magnetic Soliton Lattice on a Chiral Helimagnet. Phys. Rev. Lett. 108, 107202 (2012).

23. Togawa, Y. et al. Interlayer Magnetoresistance due to Chiral Soliton Lattice Formation in Hexagonal Chiral Magnet CrNb3S6. Phys. Rev. Lett. 111, 197204 (2013).

24. Mankovsky, S., Polesya, S., Ebert, H. & Bensch, W. Electronic and magnetic properties of 2H-NbS2 intercalated by 3d transition metals. Phys. Rev. B 94, 184430 (2016).

25. Nakayama, M., Miwa, K., Ikuta, H., Hinode, H. & Wakihara, M. Electronic Structure of Intercalation Compounds of CoxNbS2. Chem. Mater. 18, 4996–5001 (2006).

26. Zhang, C. et al. Critical behavior of intercalated quasi-van der Waals ferromagnet Fe0.26TaS2. Phys. Rev. Materials 3, 114403 (2019).

27. Tenasini, G. et al. Giant anomalous Hall effect in quasi-two-dimensional layered antiferromagnet Co1/3NbS2. Phys. Rev. Research 2, 023051 (2020).

28. Inui, A. et al. Chirality-Induced Spin-Polarized State of a Chiral Crystal CrNb3S6. Phys. Rev. Lett. 124, 166602 (2020).

29. Parkin, S. S. P., Marseglia, E. A. & Brown, P. J. Magnetic structure of Co1/3NbS2and Co1/3TaS2. J. Phys. C: Solid State Phys. 16, 2765–2778 (1983).

30. Korotin, Dm. et al. Construction and solution of a Wannier-functions based Hamiltonian in the pseudopotential plane-wave framework for strongly correlated materials. Eur. Phys. J. B 65, 91 (2008).

31. Gainutdinov, I. I., Nemudry, A. P. & Zilberberg, I. L. Using Doping to Modify the Properties of SrFeO3 and SrCoO3 Oxides: DFT Calculations of the Electronic Structure. J Struct Chem 60, 171–178 (2019).

32. Steurer, W. et al. Manipulation of the Charge State of Single Au Atoms on Insulating Multilayer Films. Phys. Rev. Lett. 114, 036801 (2015).

33. Teichmann, K. et al. Bistable Charge Configuration of Donor Systems near the GaAs(110) Surfaces. Nano Lett. 11, 3538–3542 (2011).

34. Kim, H., Chang, Y. H., Lee, S.-H., Kim, Y.-H. & Kahng, S.-J. Switching and Sensing Spin States of Co–Porphyrin in Bimolecular Reactions on Au(111) Using Scanning Tunneling Microscopy. ACS Nano 7, 9312–9317 (2013).

35. Olsson, F. E., Paavilainen, S., Persson, M., Repp, J. & Meyer, G. Multiple Charge States of Ag Atoms on Ultrathin NaCl Films. Phys. Rev. Lett. 98, 176803 (2007).

36. Hohenberg, P. & Kohn, W. Inhomogeneous Electron Gas. Phys. Rev. 136, B864–B871 (1964).

37. Kresse, G. & Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169–11186 (1996).

38. Kresse, G. & Furthmüller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Computational Materials Science 6, 15–50 (1996).

39. Kresse, G. & Hafner, J. Ab initio molecular-dynamics simulation of the liquid-metal--amorphous-semiconductor transition in germanium. Phys. Rev. B 49, 14251–14269 (1994).

40. Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 77, 3865–3868 (1996).

41. Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B 50, 17953–17979 (1994).

42. Mosey, N. J. & Carter, E. A. Ab initio evaluation of Coulomb and exchange parameters for DFT+U calculations. Phys. Rev. B 76, 155123 (2007).

43. Monkhorst, H. J. & Pack, J. D. Special points for Brillouin-zone integrations. Phys. Rev. B 13, 5188–5192 (1976).

44. Vanpoucke, D. E. P. & Brocks, G. Formation of Pt-induced Ge atomic nanowires on Pt/Ge(001): A density functional theory study. Phys. Rev. B 77, 241308 (2008).

45. Haga, T., Fujimoto, Y. & Saito, S. Electronic structure and scanning tunneling microscopy images of heterostructures consisting of graphene and carbon-doped hexagonal boron nitride layers. Phys. Rev. B 100, 125403 (2019).

There is NO Competing Interest.

Supplementary1.7.pdf
Supplementary Information

Download PDF

Version 1

posted

You are reading this latest preprint version

Tunable Single-Atomic Charge/Spin States of Intercalant Co Ions in a Transition Metal Dichalcogenide

Status:

Version 1

Abstract

Figures

Main Text

Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1