Fact-checked by Grok 2 weeks ago

Antenna effect

The antenna effect, also known as plasma-induced damage (PID), is a critical phenomenon in the fabrication of integrated circuits (ICs) where charges accumulate on isolated metal interconnects during plasma-based processing steps, such as etching and deposition, potentially leading to damage of thin gate oxides in metal-oxide-semiconductor field-effect transistors (MOSFETs).^[1]^[2] These interconnects, often long and unconnected to ground or diffusion regions, function like antennas by collecting ionized particles from the plasma, resulting in high-voltage buildup that can cause irreversible oxide breakdown.^[1]^[3] First identified in advanced semiconductor manufacturing processes, the effect poses significant challenges to IC yield and reliability, particularly as gate oxide thicknesses have scaled down to nanometers in modern nodes.^[1]^[3] The primary causes of the antenna effect stem from the inherent nature of plasma etching and related steps in IC production, including metal and polysilicon etching, photoresist ashing, ion implantation, and plasma-enhanced chemical vapor deposition.^[1] During these processes, positive charges from plasma ions accumulate preferentially on floating metal structures due to effects like shadowing in narrow spaces, where charge buildup occurs in a "latent stage" before visible damage.^[1] The severity is often quantified by the antenna ratio—the proportion of interconnect metal area to gate area—which, if excessive, amplifies the risk of charge transfer to sensitive gate oxides upon connection during later fabrication layers.^[2] This issue is exacerbated in very-large-scale integration (VLSI) designs with dense, multilayer interconnects, where long routing paths increase susceptibility.^[3] The consequences of the antenna effect include both hard failures, such as permanent transistor destruction, and soft degradations that compromise circuit performance.^[1] Damage manifests as increased gate oxide leakage currents, elevated threshold voltage variations, reduced transconductance, shortened oxide lifetime, and heightened susceptibility to noise and hot-electron effects, particularly affecting analog and mixed-signal circuits through parameter mismatches and lower gain.^[1] In severe cases, it leads to yield losses and long-term reliability failures in commercial ICs, with studies showing that conventional antenna rules may not fully account for clustered or shadowed structures, allowing latent damage to persist.^[1]^[3] To mitigate the antenna effect, IC designers employ strategies such as limiting antenna ratios through optimized routing, inserting protective diodes to provide discharge paths to the substrate, adding jumper connections to grounded diffusion regions, or reassigning interconnect layers to minimize charge buildup.^[2]^[3] Advanced tools and dynamic programming algorithms further automate layer assignment and violation fixing during physical design, ensuring compliance with foundry-specific rules while balancing area and performance overheads.^[3] These techniques have evolved with process nodes, from early aluminum metallization challenges to copper interconnects in sub-16nm technologies, underscoring the effect's ongoing relevance in semiconductor engineering.^[1]^[3]

Fundamentals

Definition and Overview

The antenna effect, also known as plasma-induced gate oxide damage (PID), refers to the unintended accumulation of electric charge on conductive interconnects during plasma-based fabrication processes in integrated circuit (IC) manufacturing, which can lead to damage in metal-oxide-semiconductor (MOS) transistors.^[1]^[4] This phenomenon arises primarily from charge buildup caused by plasma ions and electrons during key steps such as metal etching, polysilicon etching, photoresist ashing, and plasma-enhanced chemical vapor deposition (PECVD).^[1]^[4] In the context of IC fabrication, the antenna effect particularly affects thin gate oxides in advanced semiconductor nodes, such as those below 0.25 μm, where shrinking feature sizes result in dielectrics as thin as a few nanometers, making them highly susceptible to voltage stress from accumulated charges.^[5]^[4] These charges can induce high electric fields across the gate oxide, potentially causing breakdown or degradation.^[4] The relevance of the antenna effect has grown critical in modern complementary metal-oxide-semiconductor (CMOS) processes, as continued scaling exacerbates the risk due to thinner dielectrics and denser interconnects, directly impacting circuit yield, reliability, and performance.^[1]^[5] To mitigate this, design rules incorporating antenna ratios—limits on the interconnect area relative to the gate oxide area—are enforced in electronic design automation (EDA) tools during layout verification.^[1] A basic example involves long, unconnected metal lines acting as antennas that collect plasma-generated charge, which then discharges through the connected gate oxide, potentially leading to irreversible damage.^[5]^[1]

Historical Development

The antenna effect emerged as a critical concern in the late 1980s amid CMOS scaling, when gate oxide thicknesses approached 10 nm, rendering them vulnerable to plasma-induced charging during fabrication steps like etching and deposition. Initial observations linked this damage to charge buildup on floating interconnects, compromising oxide integrity and device reliability. A foundational contribution came from S. Fang and J. P. McVittie, who in 1992 developed a quantitative model and experimental validation for thin oxide damage resulting from wafer charging in magnetron plasmas, highlighting the role of plasma nonuniformity and electron shading. By the early 1990s, researchers formalized the phenomenon as the antenna effect, emphasizing how metal layers connected to gates could collect and discharge plasma-generated charges, exacerbating tunneling currents and latent defects. Seminal work by J. P. McVittie and colleagues further explored charging mechanisms using on-wafer probes, correlating plasma parameters to oxide breakdown in submicron technologies. In 1995, S. R. Nariani and C. T. Gabriel introduced differential gate antenna analysis, a wafer-level technique using test structures to monitor and predict gate oxide reliability under charging stress, enabling early detection in process development.^[6] Standardization accelerated in the mid-1990s as foundries like TSMC and Intel integrated antenna rules into design kits for 0.25 μm and 0.18 μm nodes, defining limits on antenna ratios (metal interconnect area to gate oxide area) to curb charge accumulation during multilevel metallization. These rules evolved iteratively with node shrinks, from 90 nm in the early 2000s—where copper interconnects amplified charging risks—to sub-10 nm processes by the late 2010s, incorporating cumulative ratios across metal layers.^[7] The 2000s marked milestones in advanced modeling, with studies like those by K. Krishnan et al. revealing gate area dependencies and "reverse antenna effects," where larger gates paradoxically increased vulnerability, informing refined simulation tools for yield prediction. In the 2010s, the shift to FinFET and gate-all-around architectures necessitated updated models for 3D charge flow, as fin structures altered plasma exposure and mitigation needs. SEMATECH played a pivotal role through initiatives like the SPIDER test vehicle, introduced in the 1990s and refined thereafter, standardizing metrics for process-induced damage evaluation across the industry. IEEE publications, including IEDM proceedings, further shaped consensus on antenna metrics and reliability testing.^[8] The antenna effect remains relevant in contemporary advanced nodes beyond 3 nm, where continued scaling and new fabrication techniques, such as extreme ultraviolet (EUV) lithography, introduce additional charging challenges that require evolving design and process mitigations.

Physical Mechanism

Charge Generation During Fabrication

During plasma processing in semiconductor fabrication, charge generation primarily occurs in reactive ion etching (RIE) and plasma-enhanced chemical vapor deposition (PECVD), both of which utilize radio-frequency (RF) generated plasmas to facilitate material removal or deposition. In these processes, the plasma consists of a mixture of electrons, ions, and neutral species, but charge imbalance arises because, although electrons have higher mobility and reach the wafer surface faster than slower-moving ions, the negative self-bias developed on the wafer under RF conditions suppresses the electron current while accelerating positive ions toward the surface, leading to a net positive charge buildup on exposed structures. This differential effect creates a potential difference that can drive subsequent charging phenomena.^[9]^[10] Key sources of charge in these plasmas include secondary electron emission from wafer surfaces, where incoming ions or electrons strike the surface and eject additional low-energy electrons; photoemission induced by ultraviolet (UV) light emitted from the plasma, which liberates electrons from the material; and ion bombardment, which deposits positive charge by attracting positively charged ions to negatively biased surfaces while repelling electrons. These mechanisms contribute to a net positive charging on isolated or floating structures, exacerbating the antenna effect where interconnects act as collectors. Ion bombardment, in particular, dominates in biased RF plasmas, as the self-induced negative bias on the wafer accelerates ions toward the surface.^[9]^[10] The rate of charge accumulation can be modeled simplistically as Q = (I_i - I_e) t, where Q is the accumulated charge, I_i is the ion current density, I_e is the electron current density, and t is the exposure time; this equation captures the imbalance, as I_i often exceeds I_e under biased conditions due to the wafer's negative potential, resulting in positive Q. This simplified model assumes constant currents and neglects complex sheath dynamics but illustrates the fundamental physics of imbalance.^[9] Several factors influence the extent of charge generation: higher plasma density increases both electron and ion fluxes, amplifying the imbalance; elevated bias voltage enhances ion acceleration and thus I_i, promoting greater positive charging; and gas chemistry affects emission yields, with fluorocarbon-based plasmas (e.g., CF₄ or SF₆ mixtures) exhibiting higher secondary electron yields and thus more pronounced charging compared to other chemistries like chlorine-based ones. These parameters are tuned in processes to balance etching/deposition rates while minimizing charging risks.^[9]^[10]

Antenna-Like Behavior in Interconnects

In integrated circuit fabrication, particularly during plasma etching processes, interconnect structures exhibit antenna-like behavior by collecting and conducting charge, potentially leading to gate oxide damage. Metal lines—such as polysilicon, aluminum, and copper—function as capacitors exposed to the plasma environment, accumulating charge from ions and electrons present in the etch chamber. Longer lines or those with greater surface area possess higher capacitance, enhancing their ability to collect charge efficiently as the exposed surface interacts more extensively with plasma particles.^[11] Charge collection in these interconnects is governed by the perimeter, representing the edge length directly exposed to the plasma, and the area of the metal relative to the capacitance of connected devices, which together determine the structure's susceptibility to buildup. Once accumulated, the charge flows via low-resistance paths through the interconnect network to attached gate terminals, where it can stress the thin oxide layer. The plasma's non-uniform charge distribution serves as the primary source for this collection during etching.^[12] Geometric configurations significantly influence this behavior; unconnected or floating metal segments, common in intermediate fabrication stages, amplify the effect by lacking immediate discharge routes, allowing charge to build unchecked. Vias and contacts, formed in subsequent steps, act as conduction paths that direct the charge toward sensitive nodes or provide limited discharge, depending on their connectivity at the time of exposure.^[13] A representative example involves a long, unshielded metal wire connected to a gate oxide: the collected charge Q induces a voltage V = \frac{Q}{C}, where C is the gate capacitance; if C is small relative to Q, V can surpass the oxide's dielectric strength, causing breakdown. This antenna analogy was first systematically described in studies of charge buildup during dry etching.^[11]^[14]

Design Rules

Standard Antenna Rules

Standard antenna rules in integrated circuit (IC) design aim to mitigate the risks of plasma-induced gate oxide damage by limiting charge collection on interconnects during fabrication. These rules primarily focus on geometric constraints between metal interconnects and gate structures, ensuring that the potential voltage buildup does not exceed the breakdown threshold of thin gate oxides. They are universally applied across processes to promote reliable manufacturing yields, with violations typically resolved through layout adjustments or protective elements. The core guidelines include the area rule and the perimeter rule. The area rule restricts the cumulative area of metal (or via) layers connected to a gate oxide to less than m times the underlying gate area, where m typically ranges from 10 to 100 depending on the process technology; this prevents excessive charge storage proportional to surface exposure during plasma etching. Similarly, the perimeter rule limits the total perimeter of metal lines connected to the gate to less than k times the gate area, with k commonly between 100 and 1000, as perimeter influences the edge effects of charge induction along conductor boundaries. For instance, in mature nodes like 180 nm, perimeter ratios for lower metals may reach 400:1, while area ratios for contacts or vias are often capped at 3 to 6 to account for localized charge density. To address violations of these thresholds, the diode rule mandates the insertion of reverse-biased protective diodes—typically n+ diffusion in p-substrate or p+ in n-well—on affected nets, providing low-impedance discharge paths to the substrate or well and shunting excess charge away from sensitive gates. These diodes must be placed as close as possible to the protected gate to minimize resistance in the discharge path, and their area contributes to relaxing the effective antenna ratio by a factor K (e.g., 400 for metals), often with an additional bonus for cumulative effects. Layer-specific limits impose stricter ratios on upper metal layers, such as M5 through M8 in multi-layer stacks, due to prolonged plasma exposure during their etching, which amplifies charge generation compared to lower layers like M1 or M2. Generic industry benchmarks, such as those derived from experimental validations, recommend ratios around 200:1 for poly-to-gate connections to ensure robustness across processes. Enforcement of these rules occurs through design rule checking (DRC) tools like Calibre from Siemens or Assura from Cadence, which perform layer-by-layer ratio calculations and flag violations for correction during physical verification.

Foundry-Specific Variations

Semiconductor foundries customize antenna rules to align with their proprietary processes, materials, and lithography techniques, resulting in variations that address specific charging risks during fabrication. These adaptations often involve adjusted ratio limits for metal perimeter or area relative to gate oxide, as well as layer-specific constraints to ensure reliability in advanced nodes. TSMC's rules for 7nm nodes include tighter perimeter-to-gate area ratios to address charging risks in advanced interconnects and BEOL structures. For example, these processes demand more stringent limits compared to baseline standards to account for plasma-induced effects in copper BEOL. Intel's design rules include adaptations for copper interconnects in advanced nodes to handle charge accumulation in multi-level interconnects.^[15] GlobalFoundries and Samsung incorporate flexibility in their rules, with relaxed constraints for analog blocks compared to digital sections to accommodate higher voltage tolerances and custom layouts, while integrating antenna checks into broader design-for-manufacturability (DFM) flows for yield optimization. In GlobalFoundries' GF180MCU process, for instance, metal layers enforce a maximum perimeter-to-gate area ratio of 400:1, with via ratios up to 20:1, and remedies like diode insertion or metal jumping recommended for compliance.^[16] Samsung's DFM solutions leverage closed-loop analysis to identify and resolve antenna violations alongside other manufacturability issues, enhancing integration in mixed-signal designs.^[17] By 2025, trends in 3nm and 2nm nodes involve node-specific refinements, including considerations for gate-all-around (GAA) transistors where BEOL charging from novel low-k dielectrics demands updated ratio definitions and enhanced DRC to handle denser interconnects and plasma sensitivities.^[18] These evolutions build on standard guidelines but prioritize process-unique factors like GAA channel control and backside power delivery to sustain yield in sub-3nm regimes.^[19]

Effects of Violations

Damage Mechanisms to Devices

The antenna effect leads to accumulated charge on interconnects during plasma fabrication processes, which can discharge through vulnerable device structures, primarily via Fowler-Nordheim (FN) tunneling across thin gate oxides when the induced voltage surpasses the oxide breakdown threshold. This tunneling mechanism involves quantum mechanical electron transport through the insulating oxide barrier under high electric fields, typically on the order of 5-10 MV/cm, corresponding to voltages of ~1-2 V for ultra-thin oxides around 2 nm thick, where the field strength E = V / d (with V as the voltage and d as oxide thickness) amplifies the current density exponentially. Such discharges occur because the charge collected on antenna-like metal lines, as described in interconnect behavior, seeks the path of least resistance to ground, often through the gate dielectric of adjacent transistors.^[20] The primary damage types from these discharges include soft breakdown, characterized by trapped charges within the oxide that degrade device parameters such as threshold voltage (V_{th}) by creating localized potential perturbations, and hard breakdown, which involves permanent rupture of the oxide layer leading to catastrophic short circuits.^[21] Additionally, hot carrier injection can occur in the transistor channel during high-field stressing, where energetic carriers gain sufficient energy to overcome barriers and inject into the oxide or substrate, further exacerbating interface trap generation and mobility reduction.^[22] These mechanisms are particularly pronounced in NMOS transistors due to their lower oxide tolerance compared to PMOS, as electrons in NMOS exhibit higher mobility and tunneling efficiency under positive bias.^[23] In advanced nodes, FinFET devices with high-k dielectrics face amplified risks, as the thinner equivalent oxide thickness (EOT) reduces breakdown margins, making FN tunneling more likely even with modest charge buildup.^[24] Quantitatively, the charge-to-breakdown (Q_{bd}) under constant current stress is typically on the order of 1-10 C/cm² for ultra-thin oxides, beyond which significant degradation or breakdown ensues.^[25]^[26] This corresponds to the onset of percolation paths in the oxide, linking defects and enabling irreversible damage.

Impact on Yield and Reliability

The antenna effect significantly impacts integrated circuit yield by inducing latent defects in gate oxides during plasma processing, leading to reduced manufacturing output, especially in early process development stages. Plasma-induced charging damage has been reported to decrease final product yield by 1-7% in typical fabrication environments, with extreme cases in unoptimized processes resulting in up to 50% yield loss due to increased gate oxide failures.^[27] Additionally, violations contribute to parametric shifts, such as increased variability in threshold voltage (ΔVth), which degrade transistor performance and further compound yield losses through inconsistent device characteristics.^[28] Reliability is compromised by the antenna effect through accelerated degradation mechanisms, notably time-dependent dielectric breakdown (TDDB), where accumulated charge reduces oxide lifetime and increases susceptibility to field-induced failures. In advanced technologies like 3D vertical NAND, larger antenna structures exacerbate extrinsic TDDB branches, leading to steeper failure rate increases at reduced gate oxide thicknesses. Non-compliant designs can exhibit elevated field failure rates, with charging damage correlating to higher defect densities that follow Poisson statistics for predicting overall failure probability in high-volume production.^[29]^[30] Economically, antenna violations necessitate iterative design rewrites and layout adjustments, prolonging development cycles and elevating costs in semiconductor fabrication. During the 1990s, Intel addressed potential charging issues in the Pentium processor by integrating antenna protection diodes to prevent yield-impacting oxide damage during plasma etching steps.^[31]^[32]

Mitigation Strategies

Layout and Design Techniques

Layout and design techniques address the antenna effect by modifying interconnect geometries and routing during the VLSI design process to limit charge collection on gate oxides prior to fabrication. These methods focus on breaking charge paths, providing dissipation routes, and optimizing layer usage to comply with antenna rules without altering the fabrication process itself. A primary geometric approach involves jumper insertion, where long metal wires are segmented by connecting portions via vias to intermediate metal layers that are etched later in the plasma process. This reduces the effective antenna area or perimeter connected to any gate, as charge accumulated on the upper segment dissipates during the etching of the lower layer before recombining. Optimal algorithms for jumper placement minimize the number of insertions while accounting for charge sharing across segments and obstacles in the layout, achieving solutions in linear time for Steiner trees. For instance, in complex routing trees, such methods can resolve violations with fewer than 10% additional vias compared to heuristic approaches.^[33]^[34] Another geometric fix includes adding shield layers, such as grounded metal planes or lines between signal interconnects on higher layers, to intercept and divert plasma-induced charges away from sensitive nodes. These shields, often implemented on metal layers processed after the gate oxide formation, act as barriers that ground excess charge, effectively reducing the antenna ratio for underlying structures. This technique is particularly useful in multi-layer stacks where upper metals can serve dual purposes for shielding and routing non-critical signals. Diode insertion provides an active dissipation mechanism by placing reverse-biased p-n junction diodes at the endpoints or along interconnects that risk exceeding antenna thresholds. These diodes, connected to ground or the substrate, offer a low-impedance path for accumulated charge to flow away from the gate during plasma exposure, preventing oxide breakdown. Diode sizing and placement follow rule-specific guidelines, such as inserting one diode per approximately 1000 μm of interconnect perimeter in older nodes (e.g., 180 nm) to handle typical charge densities. Polynomial-time algorithms enable simultaneous diode and jumper insertion, optimizing for minimal layout perturbation and ensuring all violations are fixed with reduced routing extensions. In practice, this can limit additional area overhead to under 5% in high-density designs.^[35]^[36] Routing strategies emphasize layer assignment to minimize exposure, such as preferring lower metal layers for connections to thin-oxide gates, where interconnect lengths are shorter and less affected by subsequent upper-layer etching. This reduces the cumulative antenna perimeter during critical fabrication steps. Incorporating via redundancy further aids mitigation by paralleling multiple vias between layers, which distributes charge flow and lowers the risk of high-current spikes through individual contacts. Advanced global routing frameworks integrate these considerations, simultaneously optimizing for antenna avoidance, via count, and timing constraints. Electronic design automation (EDA) tools automate these techniques through antenna-aware features in place-and-route flows. For example, during detailed routing, tools can automatically detect violations and insert jumpers or diodes while respecting design rules. Layer assignment modules in such systems apply optimization algorithms to prefer low-exposure paths, ensuring compliance with antenna ratios without manual fixes. These capabilities, embedded in commercial flows, significantly reduce design iteration time for large-scale ICs. In recent years (as of 2025), machine learning-integrated EDA tools have further improved automated mitigation, reducing design time for sub-3 nm processes.^[37]

Process and Material Solutions

Process tweaks during semiconductor fabrication play a critical role in minimizing charge buildup that leads to the antenna effect, particularly in plasma-based etching steps where ions and electrons can accumulate on floating interconnects. Pulsed plasma etching, which alternates between plasma-on and plasma-off periods, reduces charging damage by allowing time for charge neutralization during off-cycles, thereby lowering electron temperatures and mitigating electron shading effects that exacerbate potential differences across the wafer.^[38]^[39] Endpoint detection techniques, such as in-situ optical emission spectroscopy or interferometry, enable precise control of etch duration, shortening plasma exposure and preventing excessive charge accumulation on antenna structures.^[40] Charge neutralization methods, including electron beam irradiation or ion injection during etching, directly counteract positive charge buildup on oxide surfaces and sidewalls, reducing microtrenching and island formation while preserving etch uniformity.^[41] Material selections in fabrication flows further address antenna effect risks by enhancing charge dissipation pathways. Thicker passivation layers, such as silicon nitride or oxide overcoats, provide better shielding against plasma-induced voltages, distributing accumulated charges more evenly and lowering the risk of gate oxide breakdown in underlying devices. Alternative metals like cobalt for interconnects offer lower resistivity and better resistance to electromigration in advanced nodes.^[42] Advanced techniques integrate real-time oversight and phased processing to preempt charging. In-situ monitoring via on-chip recorders or wafer-level sensors during etching captures plasma charging dynamics, enabling dynamic adjustments to process parameters like bias power to maintain charge balance across varied antenna ratios.^[43] Multi-step deposition processes, where conductive layers are introduced incrementally to ground floating structures early in the flow, prevent prolonged exposure of unconnected metals to plasma, ensuring charge dissipation before critical oxide layers are vulnerable.^[44] These approaches, often customized by foundries, have been shown to reduce plasma-induced damage in high-aspect-ratio features.

Modeling and Verification

Simulation and Analysis Tools

Electronic design automation (EDA) tools play a crucial role in identifying and mitigating antenna effects during the physical verification stage of integrated circuit (IC) design. Synopsys IC Validator performs design rule checking (DRC) to enforce antenna rules, scanning layouts for violations such as excessive metal-to-gate area ratios that could lead to plasma-induced damage during fabrication.^[37] Similarly, Siemens EDA's Calibre nmDRC supports path-based antenna checks, enabling complex verification of charge accumulation paths in interconnects to prevent gate oxide stress.^[45] These tools integrate seamlessly with process design kits (PDKs) from foundries, incorporating technology-specific thresholds for antenna ratios to ensure compliance across nodes.^[46] Parasitic extraction tools aid in general interconnect analysis by modeling capacitances that can influence charge-related effects in layouts. Synopsys StarRC extracts resistance and capacitance (RC) parasitics with high accuracy for post-layout netlists.^[47] Calibre PEX, from Siemens EDA, generates detailed parasitic models that quantify coupling and area capacitances in dense layouts.^[48] These extraction methods provide the foundational data for timing and reliability simulations. For deeper insight into charge distribution, technology computer-aided design (TCAD) simulations model the physics of plasma-induced gate oxide damage. Synopsys Sentaurus Device simulates multidimensional charge transport and electric fields during plasma etching, predicting oxide breakdown from antenna structures by incorporating drift-diffusion models and defect generation.^[49] These simulations reveal how topographic variations in wafers exacerbate charging, allowing designers to assess damage mechanisms beyond simple rule checks.^[50] Monte Carlo methods enable probabilistic prediction of process variations in yield analysis for IC designs. These simulations generate statistical distributions of layout parameters, supporting optimization before tapeout.^[51] Key metrics in these tools center on antenna ratio calculations, defined as the ratio of interconnect area to protected gate oxide area, typically limited to 1000:1 or less per foundry PDK.^[52] DRC tools like IC Validator compute these ratios hierarchically, flagging cumulative effects across metal layers.^[37] Advances in machine learning (ML) integration into EDA flows for advanced nodes accelerate general design rule violation detection. Cerebrus and DSO.ai employ ML models to forecast DRC issues early in the design cycle, enhancing productivity. These ML integrations with PDKs enable proactive fixes.

Experimental Detection Methods

Dedicated antenna test structures are essential for evaluating the antenna effect, consisting of specialized chips that incorporate MOSFETs or capacitors with interconnects of varying geometries, such as comb-shaped, spiral, or plate-like antennas connected to thin gate oxides, to simulate different antenna ratios (AR) ranging from 100 to over 100,000.^[53] These structures allow systematic assessment of charge collection during plasma etching by measuring gate leakage current (I_g,leak) before and after the etch process; an increase in I_g,leak indicates charging-induced degradation, as the accumulated charge stresses the oxide, leading to trap generation and conduction paths.^[54] Transient fuse (T-fuse) variants in these test chips further distinguish damage from main etching versus over-etch phases by selectively interrupting charge paths post-exposure.^[53] Characterization tools provide detailed insights into oxide integrity affected by the antenna effect. Capacitance-voltage (C-V) profiling, using quasi-static (QSCV) and high-frequency (HFCV) methods at 100 kHz, detects shifts in flatband voltage (V_fb) and increased interface trap density (D_it), calculated as D_it = C_it / q where C_it is the interface trap capacitance and q is the electron charge, revealing trapped charges and interface degradation from plasma charging.^[55] Transmission electron microscopy (TEM) cross-sections visualize physical defects, such as localized oxide thinning, defect clusters, or bond breaks in the SiO2 layer, confirming charging damage mechanisms like Fowler-Nordheim injection-induced hot spots.^[22] Charge-to-breakdown (Q_bd) testing applies constant current stress (e.g., J_st = -0.25 mA/cm²) to quantify oxide endurance, where reduced Q_bd values (e.g., from 10 C/cm² in undamaged oxides to below 1 C/cm² post-plasma exposure) correlate with antenna-induced stress severity across AR variations.^[56] In-fabrication monitoring employs plasma probes to capture real-time charge dynamics during etching. Langmuir probes measure key parameters like electron density (n_e), ion flux, and plasma potential (V_p), enabling detection of non-uniform charge buildup that exacerbates the antenna effect, with typical n_e values of 10^10-10^12 cm^{-3} indicating high-risk conditions for gate oxide stressing.^[53] Wafer-level reliability tests, such as ramp voltage stress (RVS) or ramped current stress (RCS), accelerate failure detection by applying increasing bias (e.g., 3.5-7.0 V) to monitor time-to-breakdown (t_bd) or leakage onset, identifying latent antenna damage in production wafers with failure rates as low as 1% through statistical sampling.^[57] Quantitative analysis refines detection by processing experimental data for predictive insights. Statistical evaluation of failure distributions, often modeled with Weibull plots for Q_bd or t_bd data, quantifies damage variability across wafers, where shape factors (β > 1) confirm extrinsic charging defects over intrinsic wearout.^[56] In semiconductor fabs, correlation of these distributions with yield maps employs AI-driven pattern recognition, using convolutional neural networks on wafer bin maps to classify defect clusters linked to plasma charging, enabling proactive process adjustments.^[58]

References

[1]
Antenna effect (PID): Do the design rules really protect us? - EE Times
May 23, 2003 · The “antenna effect” is a common name for the effects of charge accumulation in isolated nodes of an integrated circuit during its processing.
[2]
How to stop the antenna effect from destroying your circuit - EDN
Sep 20, 2021 · To prevent the antenna effect from destroying your circuit, you need to reduce the floating metal/gate area ratio or give the charge a safe way ...
[3]
Skillfully diminishing antenna effect in layer assignment stage
Antenna effect is an important issue that critically impacts the reliability and yield of integrated circuits. The dynamic-programming-based (DP-based) ...
[4]
Thin-oxide damage from gate charging during plasma processing
The plasma-induced charge damage to small gate gate MOS capacitors is investigated by using 'antenna' structures. After an O/sub 2/ plasma step the ...
[5]
How to stop the antenna effect from destroying your circuit - EDN Asia
Sep 21, 2021 · A larger charge collection area will allow a greater charge to collect on the gate, making oxide breakdown more likely. To prevent the antenna ...
[6]
https://ieeexplore.ieee.org/document/790722
[7]
Quantitative yield and reliability projection from antenna test results
In the early 1990s, PCD became associated with the antenna effect 11,13,34 ... CMOS technology with TOX <7 nm is introduced. The proposed damage ...
[8]
https://ieeexplore.ieee.org/document/515835
[9]
https://www.sciencedirect.com/topics/engineering/plasma-damage
[10]
Physical origins of plasma damage and its process/gate area effects ...
In advanced high-k metal gate (HK/MG) technologies, plasma induced damage (PID) during process is unavoidable and has the potential to degrade device ...Missing: Marcus et
[11]
Plasma Damage - an overview | ScienceDirect Topics
Plasma damage refers to the adverse effects on materials, particularly thin films and semiconductors, caused by ion bombardment and ultraviolet radiation ...
[12]
Plasma Charging Damage
**Summary of Charge Generation Mechanisms in Plasma Processes for Semiconductor Fabrication**
[13]
A Polynomial Time Optimal Diode Insertion/Routing Algorithm for ...
This is called “antenna effect” (also called “plasma induced gate oxide damage”) [2, 3]. During met- allization, long floating interconnects act as ...
[14]
https://ssdm.jp/2021/awards.html
[15]
[PDF] (12) United States Patent - Haynes Beffel & Wolfeld LLP
Oct 23, 2009 · This destructive phenom- enon is known as "plasma induced gate oxide damage", or more colloquially the "antenna effect". The antenna effect.
[16]
Awards | SSDM2021
This is, in fact, the first paper in the world to report the so-called “antenna effect,” concerning MOS dielectric breakdown with a focus on the ratio ...<|control11|><|separator|>
[17]
[PDF] Volume 18, Issue 3, 2014 Intel® Technology Journal | 1
This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold with the understanding that ...
[18]
8.0 Antenna Ratio Rules — GlobalFoundries GF180MCU PDK 0.0.0-111-gde3240d documentation
### Summary of Antenna Ratio Rules for GlobalFoundries GF180MCU
[19]
Samsung Foundry Closed-Loop DFM Solution Leverages Mentor ...
Jun 1, 2016 · Samsung Foundry Closed-Loop Design for Manufacturability (DFM) solution leverages data from design, test and manufacturing to identify and ...
[20]
The IC designers complete guide to design rule checking
Oct 30, 2025 · If anything, the move into 3 nm, 2 nm and GAA architectures will push DRC to new limits. ... Antenna effect: Charge accumulation on interconnects ...
[21]
Impact Of GAA Transistors At 3/2nm - Semiconductor Engineering
Aug 16, 2021 · The chip industry is poised for another change in transistor structure as gate-all-around (GAA) FETs replace finFETs at 3nm and below.Missing: charging | Show results with:charging
[22]
Oxide thickness dependence of plasma charging damage
... charge-to-breakdown (Qbd), breakdown field, threshold voltage (Vth), etc ... antenna effect could occur at the wafer center8, 24. The cause of damage ...
[23]
https://www.sciencedirect.com/science/article/abs/pii/S0038110100002720
[24]
Plasma Damage in Ultra-Thin Gate Oxide Induced by Dielectric ...
A complete analysis of MOS damage induced by plasma dielectric deposition processes is presented, from the origin of the antenna effects to the consequences ...Missing: seminal | Show results with:seminal
[25]
Investigation of plasma damage effects on characteristics and ...
... antenna effect since. Experimental. The devices used in ... 1 shows the oxide thickness dependence of charge-to-breakdown of oxides used in this experiment.
[26]
Plasma‐Induced Damage on the Reliability of Hf‐Based High‐k ...
Dec 14, 2009 · This study examines the effects of plasma-induced damage (PID) on Hf-based high-k/dual metal-gates transistors processed with advanced complementary metal- ...
[27]
Charging damage to gate oxides in an O2 magnetron plasma
For exam- ple, plasma charging of floating gates can lead to significant tunneling currents in thin ... 12, 404 (1991). S. Fang and J. P. McVittie. 4872.
[28]
https://homepages.laas.fr/nolhier/ESREF2015/SESSION_B3/IP_B3.pdf
[29]
[PDF] Impacts of plasma process-induced damage on MOSFET parameter ...
During plasma exposure, devices are damaged by bombardment of high-energy ions and plasma stressing current flowing into high-k dielectric films enhanced by the ...Missing: seminal | Show results with:seminal
[30]
TDDB Lifetime Reduction From Charging Damage in a 3D Vertical ...
Apr 10, 2024 · The tests show the same intrinsic power law voltage acceleration while for larger antennas the extrinsic branches increase with reduced gate ...
[31]
Oxide damage from plasma charging: breakdown mechanism and ...
The plasma-induced charge damage to small gate area MOS capacitors is investigated by using antenna structures.Missing: impact | Show results with:impact
[32]
Antenna diodes in the Pentium processor - Ken Shirriff's blog
Several factors affect the risk of damage from the antenna effect. First, only the transistor's gate is sensitive to the induced voltage, due to the oxide layer ...
[33]
Strategies For Faster Yield Ramps On 5nm Chips
Apr 12, 2022 · Yield ramping for advanced 5nm devices or in advanced packaging, requires identification and removal of key defects at lithography through to packaging ...Missing: effect | Show results with:effect
[34]
https://ieeexplore.ieee.org/document/4305259
[35]
https://ieeexplore.ieee.org/document/4526743
[36]
https://ieeexplore.ieee.org/document/998315
[37]
https://semiwiki.com/eda/342918-checking-and-fixing-antenna-effects-in-ic-layouts/
[38]
Charging-damage-free and precise dielectric etching in pulsed
Jun 5, 2002 · Charging-damage-free φ0.05 μm SiO2 contact etching in a pulsed C2F4/CF3I plasma ... The SiO2 etching rate in C2F4/CF3I pulsed plasma did.
[39]
Mechanism of Charging Reduction in Pulsed Plasma Etching
The flux of deflected ions to the upper mask sidewalls increases enabling neutralization of the negative charge accumulated there due to the electron shading ...
[40]
In situ Metrology for Etch Endpoint Detection - Nov. 08, 2022
Nov 8, 2022 · Etch endpoint detection is used to determine when an etch process is complete and there is no more material left to etch.
[41]
Reduction of microtrenching and island formation in oxide plasma ...
Aug 6, 2025 · Electron beam irradiation during etching neutralizes positive charge buildup on the oxide island and reduces the local electric field inside ...
[42]
Plasma charging damage during contact hole etch in high-density ...
Aug 10, 2025 · The damage shows good correlation with the total exposed contact area. The safe antenna ratio is much lower than that at the conductor etch, ...
[43]
Plasma processing of low-k dielectrics - AIP Publishing
Jan 22, 2013 · One can avoid plasma damage by avoiding plasma exposure (e.g., use non-plasma deposition or cleaning methods or protect the low-k from plasma ...Missing: mitigation | Show results with:mitigation
[44]
Keeping Up Power And Performance With Cobalt
Jan 24, 2019 · Replacing tungsten contacts with cobalt has paid dividends with an approximate 60 percent improvement in contact line resistance. Replacing the ...
[45]
Wafer-Level Mapping of Plasma-Induced Charging Effect by On ...
Apr 21, 2016 · The charge stored on these in situ recorders helps to project the actual potential on the transistor gates during plasma-charging stress.
[46]
The Antenna Effect - ResearchGate
May 19, 2025 · The antenna effect is one of the primary causes of plasma and process damage, transmitting charge and energy between regions in a CMOS chip.
[47]
Reduction of plasma induced damage in an inductively coupled ...
In gate etching, pulsed plasmas have been shown to reduce notching due to mask charging. In metal etching, pulsed source power in an ECR plasma has been used ...Missing: mitigation | Show results with:mitigation
[48]
Checking and Fixing Antenna Effects in IC Layouts - SemiWiki
Mar 14, 2024 · There's something called the antenna effect that happens during chip manufacturing where plasma-induced damage (PID) can lower the reliability ...
[49]
Path-based antenna checks reduce PID susceptibility in IC designs
Designers use design rule checking (DRC) in electronic design automation (EDA) flows to find and fix patterns in the integrated circuit (IC) layout design that ...
[50]
Physical Verification: IC Validator - Synopsys
Boost productivity with Synopsys IC Validator. Achieve accurate, fast physical verification for all process nodes with seamless integration and scalability.
[51]
Synopsys StarRC - Golden Signoff Parasitic Extraction
It provides a silicon accurate and high-performance extraction solution for SoC, custom digital, analog/mixed-signal. memory IC and 3DIC designs.
[52]
Calibre xRC parasitic extraction | Siemens Software
The Calibre xRC tool provides robust parasitic extraction and accurate parasitic data for comprehensive and accurate post-layout analysis and simulation.Missing: antenna VLSI StarRC
[53]
Sentaurus Device: Multidimensional (1D/2D/3D) Device Simulator
Sentaurus Device is the most flexible and advanced platform for simulating electrical and thermal effects in a wide range of power devices such as IGBT, power ...
[54]
Impacts of plasma process-induced damage on MOSFET parameter ...
This paper presents how PID impacts on the variability and reliability characterization by focusing on two key damage creation mechanisms.
[55]
Improve Your Circuit Manufacturing Yield With Monte Carlo Analysis ...
Oct 7, 2024 · Monte Carlo statistically predicts behavior of any circuit for a set of varied component values within their tolerance range.
[56]
Predicting yield and optimizing designs by simulating thousands of ...
Jun 23, 2025 · Monte Carlo analysis simulates thousands of variations to predict how real-world manufacturing differences affect device performance. It allows ...
[57]
Antenna Effects - VLSI Concepts
Jul 19, 2008 · 'Antenna ratio' is defined as the ratio between the physical area of the conductors making up the antenna to the total gate oxide area to which the antenna is ...
[58]
CadenceLIVE Silicon Valley – OnDemand | Cadence
SoC : Sub 3nm Node Ongoing project has 15blocks planned for cerebrus exploration. Few of the blocks in initial trial showed 20% of leakage and 30% of TNS ...
[59]
Pre-Global Routing DRC Violation Prediction Using Unsupervised ...
Jun 8, 2025 · Machine learning-based design rule checking (DRC) and lithography hotspot detection can improve first-pass silicon success.
[60]
[PDF] Detection of and Protection against Plasma Charging Damage in ...
A high electric field develops across the gate and substrate of a metal-oxide- semiconductor (MOS) during plasma processing, forcing the charges through the.
[61]
A new experimental technique to evaluate the plasma induced damage at wafer level testing
### Summary of Experimental Techniques for Detecting Plasma-Induced Charging Damage
[62]
(PDF) Plasma-Induced Charging Damage of Gate Oxides
### Summary of Oxide Characterization Methods for Plasma-Induced Charging Damage
[63]
[PDF] Oxide thickness dependence of plasma charging damage
Position dependence of charge-to-breakdown (Qbd) as a function of oxide thickness and antenna area ratio. For samples with 2.6-nm-thick oxide, stress ...
[64]
Ramped current stress for fast and reliable wafer level reliability ...
A ramped dielectric stress measurement, suitable for fast wafer level reliability (fWLR) monitoring, is assessed for thin gate oxide thicknesses down to 2.2 ...
[65]
Wafer map failure pattern classification using geometric ... - Nature
May 19, 2023 · Wafer map defect pattern classification is essential in semiconductor manufacturing processes for increasing production yield and quality by ...Missing: damage | Show results with:damage