Thickness-Dependent Spintronic Terahertz Emission in MBE-Grown PtTe$_2$: From Semiconductor to Type-II Dirac Semimetal

This study demonstrates that the performance of spintronic terahertz emitters based on MBE-grown PtTe$_2$ can be optimized by exploiting thickness-driven electronic phase transitions, where a peak emission six times stronger than platinum is achieved at 10 monolayers due to enhanced spin-to-charge conversion from developing type-II Dirac band structures and interfacial Rashba effects.

The Gist

The Big Idea: Tuning a Radio to Get a Stronger Signal

Imagine you have a radio that plays music (the music is the Terahertz signal, a type of invisible light used for high-speed data). Usually, the volume of this radio is fixed by the battery inside it. If you want a louder song, you have to swap the battery for a different brand.

In the world of advanced electronics (spintronics), scientists use special materials to generate these Terahertz signals. For a long time, they used a heavy metal called Platinum (Pt) as the "battery." It works well, but its volume is stuck at a certain level. You can't make it louder without changing the material entirely.

This paper introduces a new material called PtTe₂ (Platinum Telluride). The researchers discovered something amazing: you don't need to change the material to change the volume; you just need to change how thick the layer of material is.

The Experiment: Building a Layer Cake

The scientists used a high-tech oven (called Molecular Beam Epitaxy) to build a "layer cake" of PtTe₂. They were incredibly precise, adding the material one single atomic layer at a time, going from 1 layer up to 20 layers.

They paired this cake with a magnetic layer (Cobalt) and shined a laser on it. The laser makes the magnetic layer spin, which sends a "spin current" into the PtTe₂ layer. The PtTe₂ then converts this spin into an electrical signal that shoots out as a Terahertz wave.

The Results: A Rollercoaster Ride

Here is what happened as they added more layers:

1. 1 Layer (The Semiconductor): When they had just one single layer, the material acted like a semiconductor (an insulator). It was like trying to run a race on a muddy field; the signal was almost non-existent. The "volume" was off.
2. 2 to 5 Layers (The Transition): As they added a few more layers, the material suddenly changed its personality. It shifted from an insulator to a "semimetal." The signal turned on sharply, like flipping a light switch.
3. 10 Layers (The Sweet Spot): At 10 layers, the signal reached its peak. It was six times louder than the standard Platinum reference they used for comparison.
   • The Analogy: Imagine the Platinum reference is a standard flashlight. At 10 layers, the PtTe₂ is like a high-powered searchlight.
4. 20 Layers (The Decline): If they kept adding layers beyond 10, the signal actually got weaker.
   • Why? The material became too thick and metallic. It started swallowing its own signal, like a thick fog absorbing a flashlight beam before it can escape.

Why Does This Happen? (The Physics Simplified)

The paper explains that the "volume" depends on the internal structure of the material, which changes with thickness.

• The "Topological" Highway: In the thicker layers (around 10), the electrons in PtTe₂ behave like they are on a special, super-fast highway called a Type-II Dirac Semimetal. This highway has "surface states"—special lanes where electrons can zip around without getting stuck.
• The "Rashba" Effect: Because the layers are stacked on a magnetic material, the electrons get a little "spin" (a twist) as they move, thanks to an effect called Rashba splitting.
• The Combination: When the film is just the right thickness (10 layers), these special surface lanes are perfectly formed and the "spin" is strong. This creates a perfect storm for converting the magnetic spin into a strong electrical signal.

If the film is too thin, these special lanes haven't formed yet. If it's too thick, the signal gets lost inside the material before it can get out.

The Conclusion

The researchers proved that thickness is a control knob. By simply adjusting how many atomic layers they grow, they can tune the material from being a weak signal generator to a super-powerful one.

They confirmed this by using computer simulations that matched their real-world experiments perfectly. The computer showed that the "spin" builds up at the surface of the material, and this buildup gets stronger as the film gets thicker, up until the point where the film gets too thick to let the signal escape.

In short: They found a way to make a much stronger Terahertz signal by stacking a specific material to the perfect height, unlocking a "sweet spot" where the material's internal physics works at maximum efficiency.

Technical Summary

Technical Summary: Thickness-Dependent Spintronic Terahertz Emission in MBE-Grown PtTe₂

Problem Statement
Spintronic terahertz (THz) emitters (STEs) are practical broadband THz sources, yet their performance is fundamentally constrained by the spin Hall conductivity (SHC) of the nonmagnetic conversion layer. In conventional materials like Platinum (Pt), the SHC is a fixed property once the material is selected. While topological insulators offer large spin Hall angles, their high bulk resistivity limits the overall SHC and THz emission amplitude. Transition metal dichalcogenides (TMDs) offer tunability, but most metallic TMDs suffer from conductivities orders of magnitude lower than heavy metals, resulting in suboptimal THz emission. Furthermore, the potential of exploiting the thickness-driven electronic phase evolution in TMDs to optimize spin-to-charge conversion (SCC) remains largely unexplored.

Methodology
The authors employed Molecular Beam Epitaxy (MBE) to grow high-quality 1T-PtTe₂ films with single-monolayer (ML) precision on graphene/SiC and sapphire substrates, spanning a thickness range from 1 to 20 ML. The films were capped with a ferromagnetic Cobalt (Co) layer and an Aluminum (Al) capping layer to form Al/Co/PtTe₂ heterostructures.

Structural and morphological characterization was performed using:

• Reflection High-Energy Electron Diffraction (RHEED): To confirm epitaxial growth and long-range crystalline order.
• High-Angle Annular Dark-Field Scanning Transmission Electron Microscopy (HAADF-STEM) and EDX: To verify atomic layering, stoichiometry, and interface sharpness.
• Raman Spectroscopy: To identify the 1T phase and rule out secondary phases.
• Grazing Incidence X-ray Diffraction (GIXRD): To determine in-plane lattice parameters and epitaxial relationships.

THz emission properties were measured in transmission geometry using 80 fs optical pump pulses at 800 nm. The emitted THz signals were detected via electro-optic sampling in a ZnTe crystal. To isolate the spin-current-driven signal, measurements were taken under positive and negative in-plane magnetic fields to separate magnetic (spin-driven) and non-magnetic contributions.

First-principles Density Functional Theory (DFT) calculations were conducted to model the spin-resolved band structures, spin accumulation, and spin textures of PtTe₂ slabs at various thicknesses (3, 6, 9, 12, and 15 ML) to correlate experimental trends with microscopic mechanisms.

Key Results

1. Structural Quality: The MBE-grown PtTe₂ films exhibited high crystalline quality with sharp interfaces, stoichiometric composition, and a preferred epitaxial relationship (PtTe₂(010)//Gr(010)//SiC(110)). The films maintained a smooth surface morphology with RMS roughness below 0.5 nm.
2. THz Emission Performance: A 10 ML PtTe₂ film generated a THz electric field approximately six times larger than an equivalent Pt reference (where 10 ML PtTe₂ contains the same Pt atomic density as ~1.09 nm of Pt). The emission was broadband (up to ~3 THz), linear with pump power, and exhibited the characteristic polarity reversal upon magnetic field inversion, confirming a spin-current-driven mechanism.
3. Thickness-Dependent Evolution: The THz emission amplitude displayed a distinct non-monotonic behavior:
   • 1 ML: Negligible emission, consistent with the semiconducting phase of single-layer PtTe₂.
   • 2–5 ML: A sharp onset of emission coinciding with the transition to a thin-film semimetal phase.
   • ~10 ML: A peak amplitude, reaching six times that of the Pt reference.
   • >10 ML: A decline in emission at 20 ML, attributed to THz reabsorption in the increasingly metallic bulk film.
4. Mechanism Identification: The non-monotonic trend contradicts a simple bulk inverse spin Hall effect (ISHE) model, which predicts smooth saturation. Instead, the data aligns with a multi-channel SCC process involving:
   • The development of spin-momentum-locked topological surface states.
   • Thickness-dependent interfacial Rashba splitting.
   • DFT calculations confirmed that as thickness increases, surface states decouple and sharpen, and interfacial spin accumulation increases, quantitatively reproducing the experimental emission trend.

Significance and Claims
The paper establishes that spintronic THz emission in PtTe₂ is not a fixed material property but an electronically tunable parameter controlled directly by film thickness. By traversing the electronic phase diagram from a semiconducting single layer to a type-II Dirac semimetal, the authors demonstrate that the spin-to-charge conversion efficiency can be optimized by engineering the film thickness to maximize the contribution of topological surface states and interfacial Rashba effects.

The findings introduce "thickness engineering" of van der Waals semimetals as a viable route to optimizing spintronic THz emitters and spin-orbit torques in magnetic memories (SOT-MRAMs). The work highlights that the rich phase diagram of dimensionally tunable topological materials, previously studied primarily in surface physics, can be directly harnessed for device design. The quantitative agreement between DFT-calculated spin accumulation and experimental THz emission validates the physical picture that the enhancement arises from the progressive development of the type-II Dirac band structure and associated surface/interface states, rather than a simple scaling of bulk ISHE.