Molybdenum disulfide, a prototypical TMD, captures attention because of its unique crystalline structure composed of three atomic layers: a central molybdenum atom layer flanked by sulfur atoms on either side. This extreme thinness — only three atoms thick — endows it with exceptional electronic, optical, and mechanical properties advantageous for next-generation devices. However, the challenge lies in the precise removal of the top sulfur layer during device fabrication without compromising the integrity of the underlying molybdenum. The delicate balance between effective etching and structural preservation requires innovative approaches beyond conventional physical methods.

The primary technique used for etching semiconductor materials is plasma processing. Plasma, often called the fourth state of matter, consists of ionized gases with energetic ions and electrons capable of selectively dislodging atoms from a surface. This technology, extensively researched at the U.S. Department of Energy’s Princeton Plasma Physics Laboratory (PPPL), has been instrumental for decades in material processing. Its application to TMDs, however, demands unprecedented precision because the overlap between the energy required to remove the top sulfur layer and the energy threshold that damages the molybdenum layer is exceedingly narrow.

Recent computational simulations spearheaded by a team at PPPL have unveiled a chemical strategy to widen this critical energy gap, allowing for cleaner etching of the top sulfur atoms. Key to their approach is the functionalization of the TMD surface with reactive species such as oxygen or fluorine. These atoms form chemical bonds with the sulfur on the surface, modifying the etching dynamics so that the energy barrier for sulfur removal drastically decreases from approximately 30 electron volts to around 10–14 electron volts. This decrease provides a safer operational window in plasma processing, effectively reducing collateral damage to the lower molybdenum layer.

The underlying mechanism involves not brute force but a subtle chemical assist. When a plasma ion interacts with an oxygen-functionalized MoS2 surface, it triggers the formation of sulfur dioxide (SO2) molecules. These gaseous intermediates naturally detach from the surface, making the removal of sulfur energetically easier and more selective. Fluorine coatings operate on a comparable principle, creating sulfur-fluorine compounds that similarly facilitate surface cleaning. This chemical-assisted etching presents a paradigm shift from traditional plasma processing by harnessing molecular chemistry to augment physical processes.

This insight was elucidated by Yury Polyachenko, a Princeton graduate student and PPPL associate, who emphasized that the novelty lies in the material’s chemistry rather than in the brute energetic impact by plasma ions. “We are not directly breaking the bonds,” Polyachenko explained, “but rather forming intermediate products such as sulfur dioxide, which are more easily removed.” This interplay between plasma physics and surface chemistry unlocks new avenues for nanoscale precision in semiconductor manufacturing.

While the research to date establishes a foundational understanding of the mechanism, challenges remain in quantifying and minimizing unintended damage during the plasma etching process. The team cautiously notes the imperative of characterizing the extent of molecular disruption beyond the top atomic layer, which will inform process optimization. Future experiments and simulations aim to rigorously map out the delicate trade-offs to perfect the functionalization-assisted plasma etching technique.

The implications for semiconductor technology are profound. If scalable, this methodology could be applied to a range of TMDs beyond molybdenum disulfide, including variants where molybdenum is replaced by tungsten or sulfur by selenium. Such versatility promises a diversified palette of two-dimensional materials that suit targeted electronic, photonic, or quantum applications. Exploring these analogues will determine the breadth of the technique’s utility across materials science.

These advances further complement ongoing efforts to synthesize and fabricate ultra-thin, high-performance transistors that exceed silicon’s legacy. The integration of plasma physics expertise with cutting-edge computational modeling at PPPL underscores the multi-disciplinary nature of tackling modern device challenges. Moreover, the synergy between experimental precision and theoretical insight catalyzes the innovation cycle driving semiconductor evolution forward.

The research was conducted under the auspices of the U.S. Department of Energy’s Office of Science, utilizing the resources of both the National Energy Research Scientific Computing Center (NERSC) and Princeton’s high-performance computing clusters. This computational power enabled detailed simulations of atomic-scale interactions underlying the selective plasma processing. The results were recently published in the Journal of Physical Chemistry Letters, marking a significant step toward practical implementation.

As the semiconductor industry relentlessly pursues materials and processes to sustain Moore’s Law and beyond, the ability to selectively modify atomically thin layers will be instrumental. This discovery not only provides a path to more reliable and precise etching but also deepens the understanding of plasma-matter interactions at the nanoscale. The marriage of chemical functionalization and plasma technology opens exciting possibilities for the fabrication of next-generation electronics, fueling the era of ultrathin, ultra-efficient devices.

Image Credits: Yury Polyachenko / Princeton Plasma Physics Laboratory (PPPL)

Subject of Research: Transition Metal Dichalcogenides (MoS2), plasma processing, and selective atom removal techniques for semiconductor manufacturing

Article Title: Transition Metal Dichalcogenide MoS2: Oxygen and Fluorine Functionalization for Selective Plasma Processing

News Publication Date: 27-Apr-2026

Web References:
– Princeton Plasma Physics Laboratory: https://www.pppl.gov/
– U.S. Department of Energy: https://www.energy.gov/
– Journal of Physical Chemistry Letters: http://dx.doi.org/10.1021/acs.jpclett.6c00348

References:
Polyachenko, Y. et al. Transition Metal Dichalcogenide MoS2: Oxygen and Fluorine Functionalization for Selective Plasma Processing. Journal of Physical Chemistry Letters, 2026.

Keywords:
Chemistry, Physics, Plasma physics, Computers, Technology, Transition metal dichalcogenides, Molybdenum disulfide, Plasma processing, Semiconductor manufacturing, Nanoscale fabrication, Material functionalization, Surface chemistry

Silicon’s dominance in the semiconductor industry is underpinned by decades of technological refinement centered around miniaturization, yet inevitable scaling challenges—ranging from short-channel effects to heat dissipation—have spurred the search for novel materials and architectures. Two-dimensional materials such as molybdenum disulfide (MoS₂) and tungsten diselenide (WSe₂) offer compelling advantages, including atomically thin channels ideal for electrostatic control and high carrier mobilities that surpass silicon in certain contexts. Yet, integrating these materials into complementary logic circuits, essential for power-efficient and high-performance digital electronics, has remained a significant hurdle, primarily due to difficulties in wafer-scale synthesis, doping control, and contact engineering.

Addressing these challenges head-on, the team engineered a CMOS platform by heterogeneously integrating large-area n-type MoS₂ and p-type WSe₂ field-effect transistors (FETs). This heterogeneous approach leverages the unique electronic properties of each material, enabling effective complementary operation. To optimize device performance, careful scaling of channel length and the adoption of a high-κ gate dielectric material were implemented. These design strategies were crucial for tailoring threshold voltages across both n-type and p-type transistors, ensuring that the devices operate efficiently and minimize leakage currents, which are detrimental to power consumption and overall circuit stability.

A key achievement of this research lies in the demonstration of robust device characteristics at scaled dimensions. Through precise material growth protocols and meticulous postprocessing techniques, the researchers attained transistor configurations that deliver high drive currents while suppressing subthreshold leakage. This balance is vital since the drive current dictates the switching speed and processing capability, whereas leakage currents impact power efficiency—a central consideration for any scalable semiconductor technology aiming to compete with established silicon processes.

The integrated 2D CMOS circuits exhibited functional operation below 3 volts, a testament to the meticulous engineering of device interfaces and gating architectures. The realized circuits achieved switching frequencies reaching up to 25 kHz, an impressive figure constrained predominantly by parasitic capacitances inherent in the device layout and fabrication processes. Although this frequency lags behind silicon microprocessors, it represents a pioneering proof of concept that validates the viability of 2D materials in practical computational hardware.

Noteworthy is the remarkably low power consumption observed in the 2D OISC system, operating in the picowatt range, coupled with energy-per-switching-event as low as approximately 100 picojoules. Such ultra-low power characteristics are highly desirable for applications demanding energy efficiency, including wearable electronics, implantable biomedical devices, and ubiquitous sensor networks. In these contexts, the ability to perform computations without significant power overhead could enable a new class of persistent, autonomous systems.

The authors did not stop at experimental demonstration. They further developed a comprehensive SPICE-compatible BSIM-BULK model calibrated with empirical device data, including variability across multiple transistor samples. By incorporating these realistic device-to-device variations, the modeling efforts provided invaluable insights into the scalability and practical performance bounds of the 2D material-based CMOS circuits. When benchmarked against state-of-the-art silicon microelectronics, the projections suggest that while current performance does not yet rival traditional silicon solutions, continuous material and processing improvements could close this gap, heralding a new era of 2D electronics.

This study epitomizes the complex interplay between advanced materials synthesis, device physics, and circuit design. Achieving complementary operation with two distinct 2D semiconductors required overcoming numerous technical barriers, including uniform wafer-scale crystal growth, controllable doping levels, and the formation of low-resistance, thermally stable contacts. The successful co-integration of n-type MoS₂ and p-type WSe₂ FETs on a common substrate signals a critical step toward scalable manufacturing processes compatible with existing silicon fab infrastructure.

From a broader perspective, this advance invites a reevaluation of long-standing paradigms in semiconductor technology, particularly as industry demands increasingly push beyond silicon’s fundamental limits. The ability to engineer and integrate atomically thin materials at wafer scale opens exciting frontiers not only for logic electronics but also for optoelectronics, flexible devices, and sensors. The modularity of 2D materials offers tantalizing prospects for heterogeneous integration with other emerging platforms, potentially fostering hybrid architectures that leverage the best attributes of multiple material systems.

Importantly, the research highlights the practical significance of device variability and interface engineering in 2D electronics. The authors’ approach in modeling and benchmarking accounts for real-world nonidealities, which is essential to translating lab-scale breakthroughs into industrial applications. Furthermore, the adopted high-κ gate dielectric, alongside efforts in threshold voltage tuning, illustrates how traditional semiconductor engineering principles must be adapted and refined for atomically thin materials.

The prototype one instruction set computer implemented in this work embodies a minimalist yet fully functional computing architecture, which, despite its simplicity, demonstrates the essential building blocks of digital logic implemented through 2D semiconductor technology. Its successful operation at ultra-low voltages and power levels underscores the inherent advantages of 2D materials for energy-efficient electronics while serving as a scalable platform for more complex integrated circuits.

While significant challenges remain —including improving operating frequency, enhancing device uniformity, and integrating with complementary fabrication techniques—the demonstrated system forms a pivotal foundation for future explorations in 2D microelectronics. It sparks optimism within the scientific community, signifying that 2D material-based complementary circuits may soon transition from academic curiosities to practical technologies integrated into everyday electronic devices.

In summation, the work by Ghosh and colleagues not only affirms the potential of 2D materials as viable alternatives to silicon in CMOS logic but also represents a landmark in the field’s evolution. By successfully synthesizing, engineering, and integrating large-area MoS₂ and WSe₂ transistors into a functional computing architecture, they have illuminated a promising path forward for the semiconductor industry. With continued research and development, the vision of ultra-scaled, energy-efficient 2D material-based microprocessors may well materialize, radically reshaping the landscape of electronics.

Subject of Research: Complementary CMOS circuits based on two-dimensional n-type MoS₂ and p-type WSe₂ field-effect transistors for logic computing.

Article Title: A complementary two-dimensional material-based one instruction set computer.

Article References:
Ghosh, S., Zheng, Y., Rafiq, M. et al. A complementary two-dimensional material-based one instruction set computer. Nature 642, 327–335 (2025). https://doi.org/10.1038/s41586-025-08963-7

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41586-025-08963-7