🔬 applied physics

Growth of Highly Conductive PtSe2 Films Controlled by Raman Metrics for High-Frequency Photodetectors and Optoelectronic Mixers at 1.55 {\mu}m

This study demonstrates the growth of highly conductive, semimetallic PtSe2 films on sapphire via molecular beam epitaxy and optimized annealing, where Raman metrics (specifically A1g and Eg peak widths) serve as effective quality indicators for achieving record-breaking 60 GHz bandwidth photodetectors and >30 GHz optoelectronic mixers at the 1.55 μm wavelength.

Original authors: Eva Desgué, Ivan Verschueren, Marin Tharrault, Djordje Dosenovic, Ludovic Largeau, Eva Grimaldi, Delphine Pommier, Doriane Jussey, Bérangère Moreau, Dominique Carisetti, Laurent Gangloff, Patrick Plou

Published 2026-07-13

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Eva Desgué, Ivan Verschueren, Marin Tharrault, Djordje Dosenovic, Ludovic Largeau, Eva Grimaldi, Delphine Pommier, Doriane Jussey, Bérangère Moreau, Dominique Carisetti, Laurent Gangloff, Patrick Plouhinec, Naomie Messudom, Zineb Bouyid, Didier Pribat, Julien Chaste, Abdelkarim Ouerghi, Bernard Plaçais, Emmanuel Baudin, Hanako Okuno, Pierre Legagneux

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you have a super-thin, invisible sheet of material called Platinum Diselenide (PtSe2). Think of it as a microscopic, ultra-fast highway for electricity and light. Scientists have known for a while that this material is a superstar for catching infrared light (the kind used in fiber-optic internet cables) and turning it into electrical signals. But there's a catch: making this highway smooth enough for cars (electrons) to zoom through without hitting potholes has been a nightmare.

In this study, a team of researchers at Thales Research & Technology and their partners decided to fix the road. They didn't just patch it; they rebuilt it using a high-tech oven called Molecular Beam Epitaxy (MBE) on a sapphire crystal base. Their goal? To create a film so smooth and conductive that it could handle the super-fast data speeds needed for future 6G networks.

The "Goldilocks" Recipe
The team tried cooking these films at different temperatures and with different amounts of Selenium (a key ingredient). They found that if the oven was too cool or too hot, the road became bumpy and full of cracks. The sweet spot was around 520 °C. But even at the perfect temperature, the road still had some hidden bumps.

Here is where they got clever. They realized that just looking at the surface wasn't enough. They needed to check if the layers of the material were stacked perfectly straight, like a tower of pancakes, or if they were twisted and misaligned like a wobbly Jenga tower.

To check this, they used a special tool called Raman spectroscopy. Think of this like tapping a glass to hear if it's a perfect crystal or a cracked one. The glass makes a specific "ping" sound.

One sound (called the Eg peak) tells them how smooth the road is side-to-side (in-plane).
A second, often ignored sound (called the A1g peak) tells them how straight the stack of pancakes is from top to bottom (out-of-plane).

The Secret Sauce: The High-Temp "Re-Stack"
The researchers discovered that after growing the film, they needed to give it a "re-stack" treatment. They heated it up to a scorching 690 °C for 30 minutes. This wasn't just a warm-up; it was a full structural reset.

Before this heat treatment, the material was a bit of a mess. Inside the film, the tiny crystal domains were twisted on top of each other. Imagine a stack of papers where every other page is rotated slightly; the whole stack is wobbly. This "twisted" state made the material act like a semiconductor (slower, like a dirt road) in some spots and a semimetal (fast, like a highway) in others. The result? A sheet conductance (how well it carries electricity) of only 0.5 to 0.6 mS.

After the high-temperature annealing, the magic happened. The heat allowed the atoms to shuffle around and fix their alignment. The twisted layers straightened out, forming perfect, vertical columns of crystal that went all the way through the film. The "wobbly Jenga tower" became a solid, straight skyscraper.

The Results: A Record-Breaking Highway
This structural fix changed everything. The "A1g" sound became much sharper, proving the layers were now perfectly aligned. Because the material was now a consistent semimetal all the way through, the electricity could flow freely.

The sheet conductance jumped from 0.5–0.6 mS to a record-breaking 1.6 mS for a grown film.
The team confirmed this wasn't just a fluke by checking the films after 1.5 years in the air, and they were still just as good. The material is incredibly stable.

Putting It to the Test: The 60 GHz Speed Run
To prove this new highway could handle real-world traffic, they built devices on a 2-inch sapphire wafer (a standard size for making chips). They tested two things:

Photodetectors: These are devices that catch light and turn it into a signal. They tested it with light at 1.55 µm (the standard wavelength for internet). The result? The device could respond to signals changing 60 billion times a second (60 GHz). That is a record speed for this type of material.
Optoelectronic Mixers: These are devices that mix radio signals with light signals, essential for sending data wirelessly. They built the first-ever PtSe2 mixer that works above 30 GHz.

What They Ruled Out
The paper is very clear about what doesn't work or isn't the main story here:

They explicitly argue against the idea that just looking at the side-to-side smoothness (the Eg peak) is enough. You can have a smooth surface but a twisted stack inside, and that ruins the performance.
They show that simply growing the film isn't enough; without that high-temperature "re-stack" (annealing), the material stays in a lower-quality state with lower conductivity.
They note that while graphene (another famous 2D material) is fast, it doesn't absorb infrared light as well as PtSe2. PtSe2 is the better candidate for these specific infrared applications.

The Bottom Line
The authors didn't just guess; they measured everything. They used X-rays, electron microscopes, and electrical tests to prove that the high-temperature annealing fixed the vertical alignment of the crystals. This alignment is what turned a mediocre conductor into a record-breaking one.

They suggest that if they can make the crystals even bigger and more perfect (like the ones found in nature that are peeled off rocks), the speed could go even higher. But for now, they have proven that by using a specific recipe of heat and selenium, and by checking both the "side" and "top" sounds of the material, they can build a super-fast, stable, and highly conductive material ready for the next generation of internet and wireless tech.

Technical Summary: Growth of Highly Conductive PtSe2 Films Controlled by Raman Metrics for High-Frequency Photodetectors and Optoelectronic Mixers at 1.55 µm

Problem Statement
Two-dimensional platinum diselenide (PtSe2) is a promising material for infrared (IR) optoelectronics due to its tunable bandgap, high carrier mobility, broadband absorption, and air stability. While thick (semimetallic) PtSe2 is theoretically ideal for ultrafast optoelectronics at the 1.55 µm telecom wavelength, achieving high-crystalline-quality films on low-cost, insulating substrates remains a significant challenge. Previous studies have reported carrier mobilities significantly lower than theoretical predictions, often attributed to polycrystalline nature, stacking polymorphism, and selenium vacancies. Furthermore, existing methods for evaluating film quality, primarily relying on the in-plane Raman $E_g$ peak width, have been insufficient for correlating structural properties with the high sheet conductances required for efficient optoelectronic mixers (OEMs) and photodetectors.

Methodology
The authors employed Molecular Beam Epitaxy (MBE) to synthesize thick (12–16 monolayer) semimetallic PtSe2 films on c-plane sapphire substrates. The study utilized a multi-faceted characterization approach:

Synthesis Optimization: Growth temperature ( $T_g$ ) and Selenium (Se) flux were varied. A critical step involved a post-growth annealing process at 690 °C under Se flux to improve crystallinity.
Raman Spectroscopy: Both the in-plane $E_g$ mode (~~178 cm⁻¹) and the out-of-plane $A_{1g}$ mode (~~206.5 cm⁻¹) were monitored. The Full Width at Half Maximum (FWHM) of these peaks was used as a primary metric for crystalline quality.
Structural Analysis: In-depth structural characterization was performed using Grazing Incidence X-ray Diffraction (GIXRD), High-Resolution X-ray Diffraction (HRXRD), and Scanning Transmission Electron Microscopy (STEM), including 4D-STEM diffraction mapping. These techniques analyzed in-plane domain sizes, lattice strain, interlayer distances, and the vertical alignment of domains.
Electrical and Optoelectronic Testing: Sheet conductance was measured via four-point probe and Hall-effect methods. High-frequency devices (photodetectors and mixers) were fabricated on a 2-inch sapphire wafer using coplanar waveguides (CPW) and tested at 1.55 µm.

Key Contributions and Results

Raman Metrics as Quality Indicators: The study establishes that monitoring the $A_{1g}$ peak width is as critical as the $E_g$ peak width. While $E_g$ FWHM correlates with in-plane crystalline quality, the $A_{1g}$ FWHM serves as a robust metric for out-of-plane crystalline quality. Films with narrower peaks in both modes exhibit superior structural order and higher sheet conductance.
Impact of Post-Growth Annealing: A post-growth annealing step at 690 °C was found to be essential. It induced a three-dimensional recrystallization, significantly reducing out-of-plane disorder.
- Structural Improvement: Annealing reduced the standard deviation of the local c-plane tilt angle ( $SD_{tilt}$ ) from 1.65° (as-grown) to 0.92° (annealed). It also reduced the interlayer distance and eliminated the superposition of twisted domains, resulting in vertically single-crystalline domains that span the full film thickness.
- Electrical Improvement: This structural reorganization led to record sheet conductances for grown films, reaching 1.6 mS (compared to 0.5–0.6 mS for as-grown films). The highest conductance was achieved with a Se flux of 0.5 Å·s⁻¹.
Domain Arrangement and Conductivity: The authors demonstrated that high sheet conductance requires the constituent domains to be vertically aligned and equal in thickness to the total film. As-grown films often contained superimposed twisted domains of varying thicknesses, leading to local semiconducting behavior and lower overall conductance. Annealing ensured the domains were fully semimetallic throughout the film thickness.
Device Performance: Using optimized 14–15 ML films synthesized on a 2-inch sapphire substrate, the authors fabricated:
- Photodetectors: Achieving a record 60 GHz bandwidth for a TMD-based photodetector at 1.55 µm.
- Optoelectronic Mixers (OEMs): Demonstrating the first PtSe2-based OEM with a bandwidth exceeding 30 GHz, with a conversion efficiency of -73 dB, outperforming previous graphene-based mixers at similar frequencies.
Stability: The films exhibited excellent air stability, with Raman spectra, Se/Pt ratios, and sheet conductance remaining unchanged after 1.5 years of exposure to air.

Significance
The paper claims that the simultaneous monitoring of $E_g$ and $A_{1g}$ Raman peak widths provides a rapid, non-destructive, and effective method for evaluating the overall crystallinity and predicting the electrical performance of thick PtSe2 films. By optimizing the MBE growth and implementing high-temperature post-growth annealing, the authors successfully produced high-quality semimetallic films on insulating substrates. These films bridge the gap between theoretical potential and practical application, enabling the fabrication of high-frequency optoelectronic devices operating at telecom wavelengths. The work suggests that PtSe2 is a highly promising candidate for next-generation 6G and high-speed wireless communication technologies, offering a viable alternative to graphene for IR optoelectronic mixing due to its superior IR absorption and air stability.

More like this