In the relentless pursuit of more reliable and efficient electronics, the control of metal diffusion in interconnect structures stands as a paramount challenge. Recent computational studies unveil groundbreaking insights into the diffusion behavior of copper (Cu) within tungsten disulfide (WS₂) films, potentially redefining the strategies employed in barrier layer design for semiconductor devices. These findings, derived from advanced climbing image-nudged elastic band (CI-NEB) simulations, dissect the intricate pathways that Cu atoms traverse through both ideal and defective WS₂ layers, illuminating the atomic-scale mechanisms that either hinder or facilitate their migration.
At the heart of this investigation lies the energy landscape governing Cu diffusion—a crucial factor dictating the effectiveness of WS₂ as a diffusion barrier. The computational models reveal that Cu diffusion through an ideal, defect-free WS₂ bilayer encounters an imposing energy barrier of approximately 5.6 electronvolts (eV). This barrier represents the difference between the highest energy peak and the lowest energy trough encountered by Cu atoms as they attempt to move across the tightly bonded WS₂ lattice. Such a formidable energy obstacle implies that pristine WS₂ layers are inherently robust against Cu intrusion, affirming previous suppositions about their utility in microelectronic applications.
Intriguingly, the energy required for Cu atoms to diffuse between stacked ideal WS₂ layers—a process occurring about 15 angstroms along the computed pathway—is substantially lower, measured at merely 0.5 eV. This stark contrast reflects the comparatively weaker interlayer interactions within WS₂’s layered structure, a factor that might encourage interlayer migration if not mitigated by other structural features. The study’s computational framework, visualized through detailed figures created with the VESTA software, elegantly captures these subtleties, providing clear visualizations of atomic positions and energy profiles.
Delving into the realm of imperfections, researchers introduced atomic vacancies—specifically missing tungsten (W) and sulfur (S) atoms—to simulate grain boundaries commonly found in multilayer WS₂ films. These defects serve as proxies for real-world material irregularities that often act as conduits for metal diffusion. Remarkably, Cu atoms exhibit a strong affinity for W vacancy sites, binding with an adsorption energy of -4.5 eV. This negative value signifies that Cu atoms are energetically stabilized when trapped at these defect sites, effectively creating localized energy wells that can impede further diffusion without significant energy investment.
When such defects align perfectly across WS₂ layers, forming continuous grain boundary channels, the Cu diffusion behavior alters dramatically. In this configuration, Cu atoms require a low energy input of only about 0.8 eV to transition from one WS₂ layer to the next. This significantly reduced diffusion barrier suggests that aligned defect pathways might serve as “fast lanes” for Cu migration, potentially compromising barrier performance if left unchecked. Nevertheless, the energy landscape along these pathways features a single maximum barrier of around 4.5 eV, indicating intermittent but substantial resistance points that maintain a degree of diffusion suppression.
Conversely, in bilayer WS₂ systems where defects are misaligned across layers—mimicking the grain misorientation frequently observed in practical films—the diffusion hurdles intensify. Here, Cu must overcome a much larger energy barrier of 7.8 eV to penetrate the defect-free lower WS₂ layer, alongside a 1.7 eV barrier for interlayer migration. Such high energy constraints essentially deter Cu atoms from exploiting defect sites to penetrate deeper into the WS₂ structure, underscoring the protective advantage conferred by random grain orientations.
Traditional diffusion barrier materials often rely on polycrystalline films, where the prevalence of grain boundaries and defects typically accelerates metal migration. This study disrupts that paradigm by highlighting how multilayer, randomly misaligned WS₂ films inherently thwart Cu diffusion due to the diminished likelihood of defect pathway superposition. The unique stacking and defect distribution in these films create an energy landscape that is resistant to the formation of continuous diffusion channels, thereby enhancing barrier effectiveness.
These computational insights parallel experimental observations where the thickness scaling of ALD-grown WS₂ films correlates with enhanced resistance to Cu diffusion under thermal and electrical stress. The interplay between defect alignment, grain boundaries, and nanoscale film architecture emerges as a decisive factor in designing next-generation diffusion barriers. Notably, the findings advocate for a controlled engineering of grain orientations in WS₂ films to optimize their bifunctional performance as both diffusion barriers and liners in advanced interconnect technologies.
Beyond immediate practical implications, the study enriches fundamental understanding of atomic-scale transport phenomena in two-dimensional layered transition metal dichalcogenides (TMDs). The revealed mechanisms pave the way for innovative approaches that manipulate defect landscapes and layer stacking sequences to tailor material properties at the nanoscale. Such precision could unlock new functionalities in nanodevices, potentially extending beyond interconnect technology into catalysis, sensing, and energy storage applications where controlled ion transport is crucial.
Moreover, the methodological framework employed—leveraging CI-NEB simulations to probe complex diffusion pathways—exemplifies the growing synergy between computational modeling and materials science. This approach allows researchers to predict material behavior under conditions challenging to replicate experimentally, accelerating the discovery and optimization of materials with bespoke characteristics. The detailed atomistic energy profiles generated herein offer an invaluable reference for future computational and experimental collaborations aiming to refine the atomic engineering of TMDs.
Importantly, the dependence of diffusion barriers on defect alignment within multilayer films highlights a nuanced design parameter seldom exploited in barrier technology. By fostering random grain misalignment, manufacturers might harness intrinsic material properties to enhance device reliability without resorting to thicker or more complex barrier assemblies. This strategy aligns well with the ongoing industry push towards ultrathin, low-temperature deposition methods that minimize processing costs and thermal budgets.
Furthermore, the strong Cu adsorption at W vacancies raises intriguing questions about defect passivation and chemical modification. If such defect sites can be selectively addressed or manipulated through doping, annealing, or surface treatments, it might be possible to further elevate the diffusion barriers or modulate Cu trapping behavior beneficially. These considerations open fertile ground for experimental validation and the development of complementary process steps in device fabrication.
As electronics continue to shrink and interconnect geometries become increasingly demanding, the necessity for reliable, high-performance diffusion barriers intensifies. This study’s elucidation of the atomic-scale diffusion energetics in WS₂ films not only advances fundamental science but also equips engineers with actionable knowledge to innovate barrier materials that meet future technological challenges. The balance struck between defect engineering, film morphology, and intrinsic material properties sets a new benchmark for interconnect barrier research.
In sum, the confluence of computational rigor and material insight showcased in this work heralds a new era in the design of layered TMD-based diffusion barriers. By appreciating the subtle interplay of defect chemistry and multilayer architecture, researchers can now envision smarter, more resilient materials that extend the lifespan and performance of semiconductor devices. This progress marks a pivotal step towards embedding two-dimensional materials more deeply into the semiconductor manufacturing ecosystem, leveraging their unique properties for next-generation electronics.
Looking ahead, further investigations might explore the impact of external stimuli such as mechanical strain, electric fields, or environmental factors on Cu diffusion dynamics in WS₂. Such multidimensional analyses would further refine our understanding and facilitate the tailored design of barrier systems under operational conditions. As the body of knowledge expands, the integration of WS₂ and related TMD materials into mainstream technology becomes ever more attainable, promising devices that are not only smaller and faster but also remarkably durable.
The integration of these findings into practical device fabrication could significantly mitigate the long-standing issues of electromigration and metal ion contamination. With Cu’s pivotal role in modern microelectronics, securing its containment within designated pathways enhances device reliability, reduces failure rates, and supports the relentless drive toward more powerful and compact integrated circuits. The strategic manipulation of WS₂ films, informed by these computational revelations, positions the semiconductor industry on the cusp of transformative advancements in interconnect aging and failure prevention.
Subject of Research: Copper diffusion barriers in tungsten disulfide (WS₂) films for semiconductor interconnect applications.
Article Title: Low-temperature wafer-scale growth of ultrathin tungsten disulfide for bifunctional interconnect barriers and liners.
Article References:
Mangattuchali, M.J., Astier, H.P., Chung, JY. et al. Low-temperature wafer-scale growth of ultrathin tungsten disulfide for bifunctional interconnect barriers and liners. Nat Electron (2026). https://doi.org/10.1038/s41928-026-01592-6
Image Credits: AI Generated
