JEOL USA Blog

How E-Beam Lithography Shapes the Semiconductor World

While photolithography is the engine of high-volume semiconductor manufacturing, there are certain contexts where its mask-based approach isn't feasible or precise enough. For advanced research, low-volume production, photolithography process development, and device prototyping, specialized or niche applications where flexibility and precision are paramount, the industry relies on E-Beam Lithography (EBL).

EBL is a direct-write technique that uses a finely focused beam of electrons to draw custom patterns directly onto a resist-coated surface, without the need for a physical mask. This method provides the ultimate in resolution and flexibility, making it indispensable for patterning next-generation devices. Success depends on the stability of the electron-optical column, high acceleration voltage for reduced forward scattering, and sophisticated dynamic corrections to maintain pattern fidelity across the writing area.

This is the domain of JEOL’s JBX series, the industry benchmark for high-precision, direct-write E-beam lithography.

JEOL’s Spot-Beam Vector-Scan Approach

JEOL E-beam lithography systems utilize a spot-beam vector-scan writing strategy. Unlike raster scanning, which scans the entire field line by line, the vector-scan method directs the beam only to the areas where the pattern needs to be written. High-speed electrostatic beam deflection and blanking exposes features within each write field, while step-and-repeat stage movement allows the beam to pattern the entire substrate. This approach is highly efficient for patterns with low to medium density.

Our flagship systems are purpose-built for the demands of advanced nanofabrication:

JBX-8100FS: Developed with a dual objective lens design, this system supports two operating modes, high-resolution and high-throughput, allowing users to optimize performance for either ultimate pattern fidelity or faster writing speeds. It is widely used in university nanofabs, national laboratories, and R&D lines. With acceleration voltages up to 200 kV, the system minimizes electron scattering in the resist to produce extremely sharp features, while the vector scan engine operates at up to 200 MHz. Combined with exceptional stage precision and automated functions, the system delivers reliable high-precision overlay and field stitching.
JBX-A9: Designed for leading-edge low- to medium-volume automated production and demanding R&D environments, the JBX-A9 uses a single objective lens architecture optimized to deliver high resolution and high throughput simultaneously. Together with precision stage control and dynamic correction, it achieves excellent stitching and overlay performance across large writing areas. Advanced dynamic deflection correction also ensures the beam remains precisely focused and shaped, even at the edges of the writing field, maintaining pattern integrity across the substrate.

Controlling Resolution, Proximity, and Dose

Achieving nanometer-scale patterns requires more than just a fine beam.

A spot beam system is critical for reaching the highest resolution because its small probe diameter and optimized electron optics minimize aberrations. This allows for the creation of incredibly sharp lines and features. We've demonstrated the ability to pattern sub-10 nm lines, pushing the boundaries of fabrication.

However, as electrons enter the resist and substrate, they scatter, exposing an area beyond the incident beam. As a result, the total dose at any location depends on the local pattern density — a phenomenon known as the proximity effect. Without correction, dense areas are overexposed, while sparse areas are underexposed, degrading pattern fidelity. JEOL systems integrate with third-party Proximity Effect Correction (PEC) software that modulates the dose across the pattern, ensuring every feature—from isolated lines to dense arrays—is developed with the intended dimensions.

Industrial Use Cases for E-Beam Lithography

EBL's precision makes it essential for several high-value applications in the semiconductor industry:

Photomasks: EBL’s direct-write capability is ideal for creating master photomasks used in EUV photolithography, especially those with complex Optical Proximity Correction (OPC) features and test structures. JEOL’s Variable Shaped Beam (VSB) EBL tools, like the JBX-3200MVS, are particularly well-suited for this, dramatically increasing throughput compared to spot beam tools.
Device Prototyping and Production: EBL enables direct-write fabrication of novel devices, allowing for immediate design iterations that are difficult and costly in mask-based production. This includes next-generation transistors, quantum computing components, superconducting circuits, and photonics.
T-Gate Fabrication for Radio Frequency (RF) Devices: EBL’s high resolution and precise dose control enable multi-layer resist processing used to fabricate ultra-narrow T-gates for advanced RF and high-frequency transistors. These structures are difficult to achieve with conventional photolithography and are critical for devices requiring extremely small gate lengths and precise profile control, such as high electron mobility transistors (HEMTs) used in RF amplifiers and millimeter wave (mmWave) communication systems.
Grayscale and 2.5-D Lithography: EBL can perform grayscale lithography through modulating the electron dose across a pattern. By varying the local exposure dose, grayscale lithography produces controlled differences in resist height after development, enabling the fabrication of three-dimensional or “2.5-D” structures used in micro-optics, photonics, and other advanced nanostructures.
Gratings and Periodic Structures: EBL’s high positional accuracy is crucial for fabricating gratings and other periodic structures with tight placement and periodicity tolerances. These structures are critical in applications such as optical devices, metasurfaces, waveguides, and quantum photonics.
Precision molds: Nanoimprint Lithography (NIL) requires a flawless master template, and EBL provides the precision and resolution to create these “parent stamps”. Whether for photonics, biosensors, or flexible electronics, EBL ensures high-fidelity replication, enabling low-cost mass production of complex nanoscale devices.

The Pattern-to-Proof Workflow with JEOL

A key advantage of the JEOL ecosystem is the seamless integration between patterning and verification. A typical workflow shortens the cycle of learning for process development:

Exposure: The pattern is written on a JBX series E-beam lithography system.
Develop: The e-beam resist is developed using standard cleanroom processes.
Verification: The resulting pattern is inspected on a JEOL Field Emission SEM, like the JSM-IT810. This provides immediate, high-resolution feedback on critical dimensions (CD), line-edge roughness, and overall pattern fidelity.
Optimization: The SEM data can be used to refine the Proximity Effect Correction (PEC) tables and dose strategy for the next writing run.
Atomic-Scale Confirmation: For the most advanced structures where atomic-level confirmation is required, a cross-section can be prepared and analyzed in a JEOL JEM-ARM300F Aberration-Corrected STEM.

Specifications

In E-beam lithography, performance is defined by several key metrics that determine pattern resolution, placement accuracy, and writing stability. The JEOL JBX series delivers strong performance across these parameters:

Acceleration Voltage: The JEOL JBX-8100FS and related JBX systems support acceleration voltages up to 200 kV, which reduces forward electron scattering in the resist and delivers sharper, higher-resolution patterning compared to lower-voltage systems.
Resolution: The smallest achievable feature size that can be patterned with high fidelity on a substrate. JEOL’s spot-beam tools can realize line widths smaller than 8 nm, made possible by a minimum beam diameter of approximately 1.8 nm.
Overlay Accuracy: The ability to align a new pattern layer precisely on top of an existing one. The JBX-8100FS delivers an overlay accuracy of ±9 nm in its high-resolution mode.
Field Stitching: The precision with which adjacent writing fields are joined together, crucial for generating large, seamless patterns. The JBX-8100FS offers stitching accuracy of ±9 nm in high-resolution mode.
Position Stability: Long-term beam and stage stability is critical for large-area patterning. JEOL systems maintain position stability better than 10 nm per hour, helping ensure consistent pattern placement during extended writing sessions.
Writing Speed: JEOL’s vector-scan architecture operates at up to 200 MHz, enabling efficient pattern exposure and improved throughput compared to conventional scanning approaches like Raster Scanning.

Interested in pushing the limits of nanofabrication? Contact JEOL USA to discuss transferring and refining your process recipes on the JBX-8100FS or JBX-A9, or to receive proximity effect correction (PEC) and overlap optimization support from our applications specialists.

Why Does an Electron Beam Provide Such High Resolution in Lithography?

A direct write e-beam provides high resolution in lithography because it concentrates the electron beam into the smallest possible probe size, typically just a few nanometers in diameter. This minimizes beam spreading and creates a clean, sharp exposure point. When combined with high-stability electron optics and dynamic corrections, this tiny probe allows for the direct writing of sub-10 nm features with exceptional precision, a capability perfected in JEOL’s JBX series lithography systems.

The Anatomy of Resolution in Electron Beam Lithography

Achieving the highest possible resolution isn't just about the beam. The final feature size on a wafer is a combination of several factors:

Probe Diameter: The physical size of the electron beam, determined by the system's electron optics and limited by aberrations.
Accelerating Voltage (kV): Higher accelerating voltage produces electrons with shorter wavelengths and deeper penetration into the resist and substrate. This reduces forward scattering in the resist and backscattering from the substrate, improving resolution, particularly in high-voltage lithography systems operating at 100-200 kV.
Electron Scattering: Electrons scatter as they travel through the resist and substrate (forward scatter) and bounce back (backscatter), exposing areas beyond the probe's path. This is the source of the proximity effect.
Resist Chemistry: The sensitivity and contrast of the e-beam resist determine how it responds to electron exposure. The "resist blur" can limit the final resolution.
System Stability: Mechanical and thermal stability of the stage, along with the electronic stability of the column, are critical. Any drift blurs the pattern.
Focus and Surface Height Control: Accurate control of the writing height ensures the electron beam remains properly focused across the wafer. Variations in substrate height or resist thickness can introduce defocus and degrade resolution. Integrated surface height measurement, including on transparent substrates, helps maintain proper focus during writing.
Dose Control: Precise electron dose delivery ensures that pattern features are written to their target size.

We can think of “beam blur” as the intrinsic limitation of the electron probe itself, determined primarily by the electron source and aberrations in the electron optics. This is distinct from “resist blur,” which arises from forward and backscattering of electrons within the resist and substrate. System corrections such as deflection and field calibration improve placement accuracy but do not change the intrinsic probe size.

What Makes an Electron Beam Special?

An electron beam is a circularly symmetric Gaussian beam focused to a nanometer-scale diameter. Its key advantage is the concentration of high current density into a tiny area. This produces a narrow central energy distribution, commonly described by the Point Spread Function (PSF). The PSF represents the spatial distribution of energy deposited in the resist around a single beam exposure. Probe diameter primarily determines the narrow central portion of the PSF, while the broader long-range component arises from backscattered electrons within the resist and substrate. Their combined energy distribution ultimately limits the minimum feature size that can be written.

Compared to shaped beams, which project the electron beam through a square or rectangular aperture, a Gaussian beam's narrow PSF doesn't suffer from shape convolution effects at the edges. This well-defined energy distribution maps directly to smaller minimum feature sizes and sharper corners. As demonstrated in JEOL technical reports, an electron beam with a diameter of just a few nanometers is capable of reliably patterning features well below 10 nm.

How JEOL Implementations Protect Resolution

Having a small spot is only the beginning. The entire system must be engineered to preserve that resolution during the writing process.

JBX-8100FS: This system's 100 kV accelerating voltage minimizes forward scattering, allowing the beam to penetrate the resist with less lateral spread. Its 125 MHz high-speed scan engine enables rapid, precise vector scanning. Users can select a high-resolution mode with a smaller field (e.g., 100 µm) or a high-throughput mode with a larger field (1,000 µm), with a guaranteed overlay of ±9 nm in HR mode. Optional upgrades up to 200kV accelerating voltage and 200MHz scan speed are available to further enhance capability.
JBX-A9: This system delivers high-resolution performance comparable to the JBX-8100’s high-resolution mode, while supporting the larger writing fields typically attributed to high-throughput operation. With expanded wafer handling capability and increased system automation, the JBX-A9 is designed for high-precision patterning in both research and production environments, including advanced nanofabrication facilities, quantum device research labs, semiconductor prototyping lines, and photonic device manufacturing.

Overlay and Stitching in Practice

Using a nanometer-scale probe magnifies the challenge of overlay (aligning new layers to existing patterns) and field stitching (joining adjacent writing fields). Any error in stage position or beam placement becomes a significant fraction of the feature size itself.

JEOL JBX systems manage this through a combination of a high-precision laser interferometer stage and sophisticated calibration routines that map the writing field with extreme accuracy. For final verification, we always recommend using a JEOL FE-SEM, like the JSM-IT800/IT810, for metrology. This allows you to measure the final critical dimensions (CD), line-edge roughness (LER), and overlay/stitching accuracy, providing a closed-loop process for recipe optimization.

Practical Recipe-Building Guide

To get the most out of an electron beam system, follow this practical checklist:

Probe Current: Consider the resolution that you need versus the required write time. Lower current generally means a smaller probe diameter but longer write times.
Dose Matrix: Run a dose test matrix to find the optimal exposure dose for your resist and feature size.
Proximity Correction (PEC): Apply PEC, making sure to use different parameters for dense versus isolated features.Three
Verification: After writing, inspect the pattern on a high-resolution SEM. Measure CDs and LER to check against your targets.
Iteration: Use the SEM data to fine-tune your dose and PEC tables.
Pre-flight Check: Before a critical run, allow for stage thermal stabilization, check column alignment, and run the calibration routines.

Results You Should Expect

When using a properly tuned JEOL JBX system, engineers can expect to achieve:

Sub-10-nm patterns with high-resolution resists and suitable substrates.
Overlay accuracy within single-digit nanometers when using high-resolution mode.
Excellent field stitching and CD uniformity across the entire wafer, thanks to our dynamic correction systems.

The precision of the electron beam, backed by the stability and intelligence of the JBX platform, provides the foundation for fabricating the next generation of advanced electronic and photonic devices.

Need to hit aggressive specs? Talk with JEOL about translating your target CDs and overlay requirements into a JBX recipe and SEM verification plan.

Overcoming Key Challenges in DRAM Transistor Formation Using E-Beams

As dynamic random-access memory (DRAM) technology advances through successive 10 nm-class process generations, forming uniform access transistors has become a critical challenge in advanced semiconductor fabrication. Every 1T1C memory cell contains a transistor that must perform consistently across billions of cycles. Its reliability hinges on tightly controlling the size, position, and uniformity of each transistor feature during assembly. At nanometer scales, even minor deviations in patterning, overlay, or material composition can compromise yield and retention.

Electron beams (e-beams) are able to bring a level of precision that surpasses conventional optical techniques, allowing engineers to pattern and verify transistor structures with atomic-scale control. Their extremely short wavelength and tuneable interaction with materials make them invaluable for defining, measuring, and optimizing transistor features with exceptional accuracy. In DRAM transistor formation, e-beams underpin both direct patterning and high-resolution inspection, enabling manufacturers to extend scaling beyond the limits of traditional lithography.

Why E-Beams Are Critical to DRAM Transistor Formation

Forming DRAM transistors requires precise control over geometry, alignment, and material integrity at dimensions below 20 nanometers, where conventional optical lithography and inspection systems can no longer maintain sufficient resolution and accuracy. E-beams overcome such limits by enabling the direct patterning of transistor gates and contacts with sub-10 nanometer precision, while also providing the metrology needed to measure line widths, overlay accuracy, and defect density at atomic scale. They are equally essential in photomask writing, where their nanoscale accuracy is transferred into high-volume optical production. These combined capabilities support every stage of DRAM transistor formation, encompassing gate definition, structural verification, and overall process stability, and allow manufacturers to deliver the consistency and yield demanded by advanced memory design.

Addressing DRAM Transistor Formation Challenges Through E-Beams

The Role of Photomasks in Translating E-Beam Accuracy to Production

Although e-beam lithography provides unmatched precision for research and prototype development, high-volume DRAM manufacturing depends on optical lithography guided by e-beam-written photomasks. These masks serve as the master templates that transfer nanoscale accuracy onto thousands of wafers.

E-beam writers define each photomask with nanometer precision, enabling advanced features such as optical proximity correction (OPC) and phase-shift designs that enhance the resolution of optical systems. Before production, e-beam inspection tools verify mask integrity, ensuring the intended transistor geometries and layer alignments are faithfully reproduced throughout the manufacturing process. Together, these steps connect e-beam precision with mass production, enabling the dimensional accuracy necessary for advanced DRAM devices.

Advancing DRAM Transistor Formation with JEOL USA

The challenges of DRAM transistor formation can be met through electron beam technology, which delivers the precision and process control that enable reliable device scaling. JEOL USA provides advanced electron-beam lithography systems for transistor patterning and photomask writing, high-resolution scanning electron microscopes (SEMs) for inspection and metrology, and analytical transmission electron microscope (TEM) and SEM systems with energy-dispersive X-ray spectroscopy (EDS) and electron backscatter diffraction (EBSD) capabilities that enable material characterization and process optimization in semiconductor manufacturing. Uncover more information about our e-beam and analytical technologies and how they can advance your semiconductor processes by contacting our technical experts today.

How Does Field Ionization Work?

Mass spectral interpretation begins with the molecular ion. Electron ionization (EI) has traditionally been the primary method for generating this species, valued for its reproducibility and extensive reference mass libraries. Yet the high-energy electrons at 70 eV used in EI often deposit excess internal energy, leading to extensive fragmentation and suppressing the molecular ion. Field ionization (FI) avoids such an outcome by eliminating collision-based ionization and instead extracting an electron using an intense electric field. Understanding how field ionization works reveals why it excels at preserving molecular ions while retaining meaningful structural information.

The Fundamental Physics: Quantum Mechanical Tunneling

Field ionization is governed by quantum mechanical principles rather than classical collision ionization processes. In a field ionization source, analyte molecules pass through an extremely strong electric field, typically on the order of 10⁷ to 10⁸ V/cm, established between a high-potential emitter and counter electrode. Under normal conditions, electrons remain bound to molecules through a potential energy barrier that exceeds the molecule’s ionization energy. The applied electric field distorts this barrier, effectively narrowing it on one side of the molecule. When the barrier becomes sufficiently thin, quantum mechanical tunneling can occur. This behavior reflects the Heisenberg Uncertainty Principle, which states that electrons are delocalized probability distributions instead of fixed, classical particles. In the presence of the distorted barrier, the quantum nature of the electron enables tunneling out of the molecule without a physical collision. Ionization therefore proceeds with minimal energy transfer, producing a molecular radical cation in a low vibrational state and largely preserving molecular structure.

Emitter Architecture and Field Concentration

Generating the electric fields required for field ionization depends as much on emitter geometry as on the voltage applied between the emitter and counter electrode. Electric field strength scales in proportion to the applied potential divided by the radius of curvature of the conducting surface, so reducing that radius sharply increases local field intensity. FI emitters are engineered specifically to leverage this relationship. A typical design employs a fine metal wire, often around 10 micrometers in diameter, coated with carbon dendritic structures commonly referred to as whiskers or micro-needles. The dendritic structures terminate in tips with nanometer-scale radii of curvature, forming regions of extreme electric field concentration. In these regions, field strengths readily reach levels required for ionization, enabling field ionization at voltages compatible with practical mass spectrometer operation.

Comparative Dynamics: Field Ionization and Other Soft Ionization Methods

Occupying a distinct position among soft ionization techniques, FI minimizes secondary chemical interactions during ion formation. Chemical ionization (CI), by contrast, relies on reagent gases and ion-molecule reactions that commonly introduce protonated species, adducts, and chemical background. Field ionization bypasses these pathways entirely, producing spectra typically dominated by the intact molecular ion and often simpler to interpret. Photoionization (PI) is another low-energy alternative, but its effectiveness depends on photon absorption cross sections that are often low for non-polar compounds. Because FI does not depend on photon absorption, it is suitable for saturated hydrocarbons and other chemically inert species. In addition, FI operates without elevated source temperatures, reducing thermal stress and extending its applicability to thermally labile compounds. Such a combination of advantages makes FI a valuable option for analyses that require molecular ion preservation and minimal chemical interference, such as petroleomics workflows and the analysis of thermally labile compounds.

Analytical Applications in Complex Characterization

Field ionization is applied when molecular ion preservation is required for reliable mass spectral interpretation and to provide intact molecular ions for the identification of unknown compounds. The ionization process transfers very little excess internal energy, limiting fragmentation at the point of ion formation and enabling the molecular ion to remain dominant in the resulting spectrum. Access to intact molecular ions under such conditions enables analytical workflows that depend on direct molecular weight determination and composition assignment to proceed without ambiguity introduced by fragmentation.

Saturated hydrocarbon analysis
Linear and branched alkanes often yield weak or absent molecular ion signals in EI, limiting confident molecular weight determination. Field ionization overcomes this limitation by consistently generating intact molecular ions for non-polar hydrocarbons, supporting petroleomics workflows such as group-type and compositional analysis.
High-resolution mass spectrometry synergy
Field ionization spectra retain intact molecular ions, allowing integration with high-resolution mass spectrometry. Accurate mass measurements can be applied directly to molecular ions, supporting reliable elemental composition assignments based on exact mass, even in chemically complex mixtures.
Oligomeric and polymer characterization
Low- to mid-molecular-weight polymers can be characterized through FI by preserving molecular ion series with minimal fragmentation, maintaining molecular weight distributions, and enabling direct assessment of oligomer populations.

Advancing Mass Spectrometry with JEOL USA

Field ionization enables direct control over ion formation through retaining molecular ions with minimal excess internal energy. Its capabilities enable analysts to tailor ionization behavior to the chemical problem, balancing molecular ion preservation against structural information. JEOL USA implements FI within GC-MS platforms like the JMS-T2000GC AccuTOF™ GC-Alpha 2.0 Gas Chromatograph-Time-of-Flight Mass Spectrometer and AccuTOF™ GCxGC Mass Spectrometer, as well as EI and FI within a single ion source. Access to multiple ionization modes ensures analysts can select the optimal balance between fragmentation and molecular ion preservation, extending robust molecular characterization across diverse analytical challenges. Reach out to JEOL USA for more information about our FI-enabled GC-MS solutions and available method development support.

An Introduction to FIB-SEM

The behavior of materials and biological systems is often shaped by the hidden structures that lie beneath their surfaces. A crack deep within a metal, a fault in a semiconductor device, or the architecture of a biological cell can hold insights about a sample’s structure and function that surface imaging alone cannot provide. Traditional microscopy techniques each have their place in characterizing surfaces and structures. Scanning electron microscopy (SEM) delivers detailed surface information, while transmission electron microscopy (TEM) reaches atomic resolution, although it does require ultrathin specimens. However, neither SEM nor TEM alone can easily provide subsurface or three-dimensional data. Researchers therefore have turned to focused ion beam-scanning electron microscopy (FIB-SEM), which overcomes these limitations by combining precision milling with high-resolution imaging. It has since become a well-established tool across many areas that require advanced microscopy, including materials science, semiconductors, life sciences, and energy research.

What is FIB-SEM?

A FIB-SEM is a dual-beam instrument that brings together:

A focused ion beam (FIB), which directs a stream of ions onto the sample to cut, mill, or sputter away thin layers of material.

A scanning electron microscope (SEM), which uses a beam of electrons to produce high-resolution images of the newly exposed surface.

By combining an ion beam for material removal and an electron beam for imaging, FIB-SEM helps scientists to:

Examine hidden structures beneath the specimen’s surface.
Collect a sequence of images that can be reconstructed into a 3D model.
Produce ultrathin lamellae at selected locations for TEM analysis.
Deposit protective coatings or modify surfaces as needed.

Rather than serving a single function, FIB-SEM unites fabrication and imaging in one system, making it uniquely versatile and a powerful tool for investigating complex structures.

How Does FIB-SEM Work?

The Workflow

FIB-SEM workflows are built around a repeated cycle of milling and imaging. It typically consists of four key stages:

Identify the region of interest
The SEM is used to visualize the sample surface and locate the exact area of investigation. Researchers can zoom in on features such as defects, grain boundaries, or regions of biological tissue that require closer examination.
Mill with the ion beam
A finely focused beam of ions, usually gallium, is directed at the chosen site. With the adjustment of the beam current, operators can switch between fast, coarse milling for removing bulk material and gentle, fine milling for polishing delicate features, ensuring the resulting cross-section is clean and ready for imaging.
Image with the SEM
At this point, the SEM directs an electron beam across the newly exposed surface, generating contrast that highlights the sample's surface structures and microstructural features.
Repeat the cycle
Alternating between milling and imaging allows researchers to cut through the sample layer by layer. The set of acquired scans can then be assembled into a 3D reconstruction or used to prepare an ultrathin lamella for TEM analysis.

The Components Behind the Workflow

Several key components are required to make the FIB-SEM workflow possible:

Ion source: Gallium liquid metal ion sources provide precise, stable milling.One
Electron column: This part of the SEM generates and focuses the electron beam, producing high-resolution images that reveal surface detail and compositional contrast.
Stage and chamber: A motorized, multi-axis stage positions the sample with nanometer accuracy, and large chambers accommodate varied specimen sizes.
Detectors and accessories: Options such as STEM detectors, EDS for elemental mapping, and gas injection systems (GIS) expand the analytical capabilities of the FIB-SEM and support processes like protective coating and lamella preparation.

Applications of FIB-SEM

The adaptability of FIB-SEM has made it valuable across many fields:

Materials science: Studying grain boundaries, porosity, and failure mechanisms, often using protective coatings to safeguard delicate regions of the sample during milling.
Life sciences and neuroscience: Imaging tissues and cells in 3D, mapping neural networks, and preserving ultrastructure through cryo-FIB-SEM workflows.
Energy and geosciences: Characterizing battery electrodes, porous rocks, and other complex materials to understand performance and degradation.
Semiconductors and electronics: Inspecting device defects, modifying circuits, and preparing lamellae for TEM, with gas-assisted deposition enabling both shielding and microfabrication.

Advantages of FIB-SEM

FIB-SEM offers a unique set of benefits:

High-resolution 3D imaging that reveals both surface and subsurface features, providing insights that are unavailable from conventional microscopy.
Site-specific targeting, which allows researchers to focus on regions of interest without damaging surrounding material.
Integrated milling, imaging, deposition, and analysis, reducing the need for multiple instruments and streamlining workflows.
Flexibility across scientific and industrial applications, from failure analysis to biological imaging, making it versatile enough to cater to diverse research challenges.

Advancing FIB-SEM Research With JEOL Solutions

FIB-SEM has established a means of exploring beyond surface imaging and studying subsurface structures in greater detail. The JIB-PS500i from JEOL supports these investigations with precise ion milling, seamless integration with TEM, and cryogenic options for sensitive samples. Its spacious chamber and 5-axis motorized stage accommodate diverse specimens, while integrated STEM and EDS detectors keep imaging and analysis within a single platform.
For researchers working with frozen hydrated or otherwise delicate materials, the CRYO-FIB-SEM CryoLameller adds a dedicated cryogenic workflow. Liquid-nitrogen cooling, cryocooled transfer, and controlled low-temperature milling help maintain sample integrity and enable the production of high-quality TEM lamellae. Together, the JIB-PS500i and CRYO-FIB-SEM CryoLameller provide complementary approaches for conventional and cryogenic FIB-SEM studies. Reach out to our specialists to learn which configuration would best support your research.

Using High-Resolution SEM and TEM for Advanced Semiconductor Packaging

Semiconductor packaging is undergoing significant changes. For years, the focus was on shrinking the pitch of micro-bumps and Through-Silicon Vias (TSVs). Now, driven by the insatiable demand for more bandwidth, lower power consumption, and reduced latency, the industry has adapted the latest technology of direct copper-to-copper (Cu-Cu) hybrid bonding.

This shift from micro-scale interconnects to direct, atomic-scale bonds marks a fundamental change in what we need to measure and control. The challenge is no longer just seeing the bump; it’s about characterizing the interface itself. Reliability and yield are now decided by nanometer-scale details: oxide layers, voids, contaminants, and crystal grain alignment.

To succeed in this new paradigm of heterogeneous integration, packaging engineers need a complementary analytical toolkit. JEOL's Field Emission SEM (FE-SEM) and atomic-resolution Transmission Electron Microscopy (TEM/STEM) provide the complete workflow required to characterize these critical interfaces, from the die level down to the atomic scale.

What Must Be Measured Now

In an era of hybrid bonding and high-density interconnects, the list of critical-to-quality parameters has grown longer and moved to a much smaller scale. Success depends on precise control over:

Interface Cleanliness & Planarity: Pre-bond oxide thickness and surface uniformity are paramount for a successful bond.
Post-Bond Integrity: Detecting nanometer-scale voids, gaps, and contaminants at the bond interface.
Dimensional Control: Measuring copper recess and dielectric thickness with nanometer precision.
Alignment: Verifying die-to-wafer alignment at the nanometer level.
Crystallography & Chemistry: Understanding the copper grain structure, orientation, and the presence of unwanted oxides or contamination at the bond line.
Legacy Joint Reliability: In traditional micro-bumps, identifying and quantifying Intermetallic Compound (IMC) phases that dictate long-term reliability.

JEOL SEM for High-Throughput Packaging Inspection

For rapid, high-throughput inspection of large areas, JEOL’s FE-SEM is the first line of defense.

The JSM-IT800/IT810 FE-SEM with integrated Energy Dispersive X-ray Spectroscopy (EDS) is ideal for characterizing micro-bump arrays, hybrid bond pads, and Redistribution Layers (RDLs).

Imaging Modes: Use Secondary Electron (SE) imaging for high-resolution topographical detail and Backscattered Electron (BSE) imaging for powerful materials contrast, which is perfect for distinguishing between solder, IMCs, copper, and underfill materials.
Automated Workflows: Our SEM Center software enables automated recipes for inspecting hundreds of bond pads, providing statistically significant data on pad integrity and defects.

The quality of any analysis depends on the quality of the specimen. The JEOL Cross Section Polisher™ (CP) uses a broad, low-energy argon ion beam to create large, clean, and artifact-free cross-sections of challenging multi-material stacks containing solder, copper, and polymers. Standard and cooling stage options prevent heat damage to sensitive materials, while air-isolation transfer options protect the sample from atmospheric contamination.

JEOL HR-TEM/STEM for Interfaces That Decide Reliability

When an interface fails, or when you need to certify a new process, you must go beyond the resolution limits of SEM. JEOL's aberration-corrected TEM/STEM provides undeniable, atomic-scale proof.

The JEM-ARM300F (GRAND ARM) is designed for this exact challenge.

Sub-Ångstrom Imaging: High-Angle Annular Dark-Field (HAADF) STEM imaging can easily resolve nanometer-scale oxide layers, voids, and defects directly at the Cu-Cu bond interface.
Chemical Analysis: Integrated EDS and Electron Energy Loss Spectroscopy (EELS) can perform line scans across the bond to map the precise location and concentration of oxygen, carbon, and other contaminants. EELS can even determine the oxidation state of copper, distinguishing between problematic oxides and pure metal.
Crystallographic Information: Using diffraction modes, you can analyze crystal grain orientation and measure nanoscale strain, both of which are critical for predicting bond reliability.

Workflow Examples: From Process Control to Failure Analysis

1. Verifying a Hybrid Bond Interface:

Create a cross-section of the bond area using the JEOL Cross Section Polisher™.
Use a JSM-IT800 FE-SEM for a rapid survey and EDS mapping to confirm the location and general composition.
Prep site-specific TEM lamella targeting the bond interface using JEOL dual beam PS500i.
In the JEM-ARM300F, acquire a STEM-EELS oxygen map across the interface. The resulting data can be used to set clear pass/fail criteria (e.g., oxide thickness must be <1 nm, void area fraction must be <0.5%).

2. Assessing Micro-bump Reliability:

Use the CP to create a clean cross-section through a series of solder bumps.
In the SEM, use Backscattered Electron (BSE) imaging to clearly highlight the different Intermetallic Compound (IMC) layers.
Use integrated EDS to perform phase identification and quantify the elemental composition of each IMC layer.
For advanced failure analysis, a TEM sample can be prepared to measure IMC layer thickness with nanometer precision and investigate the morphology of any micro-voids.

Reporting for Actionable Engineering Insights

A comprehensive engineering report moves beyond pretty pictures. It translates raw data into actionable metrics. Using JEOL's integrated suite, your reports can include:

Histograms of bond pad alignment error.
Distributions of oxide thickness measurements from EELS data.Three
Void size and area fraction distributions.
Quantification of IMC phase fractions and layer thicknesses.

This data allows you to set firm, statistically validated acceptance windows for your semiconductor packaging processes, directly linking nanoscale metrology to device reliability and yield.

Ready to see your interface? Set up a JEOL packaging interface study—SEM + ARM-class TEM recipe designed for your pad stack and pass/fail metrics.

How is Metrology Used in Failure Analysis?

Every credible failure analysis is supported through metrology, the science of measurement. When a component, material, or system fails, engineers need more than visual clues. They require quantitative data that reveals how and why the failure occurred. In semiconductor devices, where functionality depends on nanoscale geometries, interface integrity, and material uniformity, metrology is essential for distinguishing normal process variation from true failure mechanisms. By characterizing geometry, surface features, and internal structure with traceable accuracy, it ensures the analysis is grounded in verifiable results and stands as a reliable determination of the underlying cause of failure.

Why Measurement Quality Matters in Failure Analysis

Before engineers can test a hypothesis, they must record exactly how the part failed, as these details are what enable a failure analysis to be trustworthy. Without reliable data, conclusions are speculative and corrective actions may miss the true cause of failure. Accurate metrology turns observation into quantifiable fact, ensuring each dimension, defect, and deviation is captured with confidence.

In practice, this means:

Quantifying geometry and defects - reveals how features such as crack width, distortion, void size, and coating thickness contributed to failure.
Comparing "as-failed" to "as-designed"- identifies if dimensions have drifted beyond tolerance and where a process or material deviation occurred.
Ensuring traceability and reproducibility - provides calibrated, uncertainty-defined results that can withstand quality or warranty scrutiny.

When measurement quality is strong, engineers can trust the data enough to trace the failure back to its true cause and understand how it developed in the first place.

How Metrology Is Used in Failure Analysis

1. Initial visual and dimensional inspection

The process starts with a non-destructive evaluation of the failed part. Optical or coordinate measurements capture overall geometry, identifying distortion or dimensional drift relative to design drawings. Collecting these measurements early helps engineers preserve the "as-failed" condition before moving on to further testing.

2. Surface metrology and form measurement

Scanning electron microscopy (SEM) delivers surface metrology that reveals fine details of pattern deformation, line-edge roughness, contamination, and localized damage within semiconductor structures. Quantifying roughness and topography uncovers how stress, friction, or environmental conditions contributed to the failure.

3. Internal defect detection

Focused Ion Beam (FIB) 3D reconstruction provides high-resolution, site-specific internal metrology by sequentially milling and imaging material volumes at the nanoscale. This approach enables engineers to visualize and measure buried voids, interconnect discontinuities, delamination, and interface defects that cannot be resolved through surface inspection alone and are often responsible for electrical or reliability failures in semiconductor devices.

4. Cross-section and microstructural metrology

If destructive sectioning is needed, optical or electron microscopy measures grain size, inclusion distribution, or microcrack geometry to determine whether the cause of failure lies in material quality or processing.

5. Correlation with design and process data

Metrology measurements are compared with design tolerances and manufacturing records to locate where deviation occurred, be it in raw material, machining, or assembly.

6. Validation and modeling

Dimensional and microstructural data generated through metrology feed directly into computational models such as finite-element analysis (FEA). Using real, measured parameters, engineers can simulate stress propagation or fatigue behavior to assess how well the modeled mechanism matches the failure observed.

Throughout the investigation, metrology is able to supply consistent data that supports the failure analysis, allowing engineers to connect observation with the mechanisms that contributed to the defect and identify the improvements that can prevent it from recurring.

Techniques That Enable Quantitative Failure Analysis

Failure analysis draws on several branches of metrology, which offer unique insight into geometry, surface condition, or internal structure:

Dimensional metrology- Coordinate measuring machines (CMMs), laser trackers, and optical scanners quantify part geometry and tolerance drift.
Surface metrology- Stylus and optical profilometers measure roughness and waviness, while SEM provides high-resolution images of fracture surfaces and wear tracks.
Internal defect metrology-Internal defect metrology- X-ray and CT scanning allow for the measurement of internal voids, inclusions, or delamination without altering the sample.
Electron microscopy-based metrology- SEM and Transmission Electron Microscopy (TEM) measure micro- and nano-scale features such as inclusions and interface integrity.
Analytical extensions- Energy-dispersive X-ray spectroscopy (EDS), electron energy loss spectroscopy (EELS), and mass spectrometry add chemical information that links structure to composition.

Together, these complementary metrology techniques support evidence-based failure analysis across a range of industries, including aerospace and energy.

How Electron Microscopy Is Used in Semiconductor Failure Analysis

In the semiconductor industry, where critical features reach the nanometer scale, Transmission Electron Microscopy (TEM) and Scanning Electron Microscopy associated with Focused Ion Beam (SEM/FIB) provide the resolution to identify failures within complex device architectures, including dislocations, voids, interface defects, and other nanoscale features that influence device performance.

A typical semiconductor failure-analysis workflow includes:

Fault localization- Electrical testing or SEM identifies the suspect region.
Sample preparation- FIB milling thins the site to tens of nanometers, generating an electron-transparent lamella.
TEM or STEM imaging- High-energy electrons pass through the sample, revealing dislocations, voids, or interface defects with atomic-scale clarity.
Analytical mapping- EDS and EELS quantify elemental distribution and chemical bonding to connect material changes with electrical behavior.
Correlation- JEOL software enables quantitative metrology of critical dimensions within device features, such as interconnects, gate structures, dielectric layers, and material interfaces, allowing engineers to confirm whether voids, oxide breakdown, or interface defects caused the functional failure.

SEM, FIB, and TEM metrology supplies evidence that engineers can use to diagnose device failures accurately and make informed changes that enhance manufacturing outcomes.

Strengthening Failure Analysis with Metrology Solutions from JEOL USA

Modern failure analysis requires precise metrology tools that can measure and resolve defects and variations that lead to component or system issues, including nanoscale dimensional and material variations common in semiconductor devices. JEOL USA offers advanced electron microscopy, mass spectrometry, and electrical analysis systems that can characterize structural features, identify chemical contaminants, and locate electrical defects with high accuracy. Speak with JEOL USA to learn more about our available products, including our TEMs and SEMs, designed for high-precision failure analysis and how they can help drive comprehensive, data-driven investigations that accelerate root-cause identification and support more resilient product development cycles.

What is a FinFET and How Does it Work?

For decades, the planar MOSFET was the bedrock of integrated circuits. But as semiconductor nodes shrank, the physical limits of this 2D design began to show. Engineers faced a critical challenge: "short-channel effects." As the distance between the source and drain decreased, the gate lost electrostatic control, leading to current leakage and increased power consumption.

The industry needed a new transistor geometry. The solution was to go 3D.

Enter the FinFET. This multi-gate transistor architecture revolutionized semiconductor design by enabling the development of advanced nodes. Instead of a planar channel, the FinFET uses a vertical silicon "fin" that rises from the substrate. The gate is wrapped around this fin on three sides, providing superior electrostatic control and mitigating the short-channel effects that plagued its planar predecessors.

This "multi-gate" approach is a key concept. A multi-gate device, as the name implies, has more than one gate on a single transistor. This family of devices includes not only FinFETs but also future architectures like Gate-All-Around (GAA) FETs, which promise even greater control and scaling.

A Quick Primer: Planar vs. FinFET

To understand the difference, imagine a simple cross-section:

Planar MOSFET: A flat, 2D channel lies on the silicon substrate. The gate sits on top, controlling the flow of current.
FinFET: A 3D fin of silicon rises vertically. The gate wraps around the fin, creating a larger, more effective control area.

Key terms to know for FinFETs include:

Fin Height (Hfin): The height of the silicon fin.
Fin Width (Wfin): The thickness of the fin.
Pitch: The distance between adjacent fins.
Gate Length: The length of the gate as it wraps around the fin.

How a FinFET Actually Operates

The genius of the FinFET lies in its 3D gate control. By wrapping the gate around the fin, the FinFET achieves:

Superior Depletion: The gate can more effectively deplete the channel of charge carriers, leading to a much lower "off-state" leakage current.
Steeper Subthreshold Slope: The subthreshold slope is a measure of how quickly a transistor can switch from "off" to "on." A steeper slope means a more efficient switch, and FinFETs excel in this regard. This translates to lower power consumption.
Increased Drive Current: At the same supply voltage (Vdd), a FinFET can deliver a higher drive current than a planar device. This means faster, more powerful processors.
Parasitics to Watch: The 3D structure also introduces new parasitic capacitances and resistances that must be carefully managed during design and characterization.

Boxed Math: The Fin Aspect Ratio

The electrostatic control of a FinFET is directly related to its fin aspect ratio (Hfin/Wfin). A taller, thinner fin (a higher aspect ratio) provides better gate control and reduces short-channel effects. However, it also presents significant manufacturing and characterization challenges.

Complementary FET (CFET) architectures represent the next step beyond FINFET-based scaling. In a CFET, the n-type and p-type transistors that form a completely metal-oxide semiconductor (CMOS) pair are vertically stacked rather than placed side by side. This vertical arrangement significantly reduces standard cell area without relying on further lateral pitch scaling. CFETs build directly on the gate-control concepts established by “Gate-All-Around” devices, but shift the primary scaling benefit toward stacking and layout efficiency, making them a leading candidate for technology nodes below 2 nm.

Why 3D Geometry Demands New Characterization

The complex, three-dimensional nature of FinFETs presents new challenges for process control and failure analysis. Simply put, you can't characterize what you can't see. Key challenges include:

Targeting Specific Fins: In dense arrays of fins, isolating a single, specific fin for analysis is a major hurdle.
Preserving Interfaces: The interfaces between the high-k dielectric and the metal gate (HKMG) stack are critical to device performance. Preparing a sample for analysis without damaging these delicate layers is essential.

This is where JEOL's purpose-built workflows come in.

The JEOL Workflow: From Specimen to Atomic-Scale Analysis

JEOL provides an end-to-end solution for FinFET characterization, from specimen preparation to atomic-scale imaging and analysis.

Site-Specific Preparation:
- FIB-SEM: Our Focused Ion Beam (FIB) and Scanning Electron Microscope (SEM) systems allow for precise, site-specific milling to isolate the fin of interest.
- Cross Section Polisher™ (CP): For wide, clean, and damage-free cross sections, our CP tools use a broad argon ion beam to gently polish the sample surface. For sensitive materials, our air-isolation options protect the sample from atmospheric contamination.
High-Resolution Imaging and Analysis:
- JSM-IT800/IT810 FE-SEM: These Field Emission SEMs provide ultra-high-resolution imaging for critical dimension (CD) measurements, line-edge roughness analysis, and defect localization. Integrated Energy Dispersive X-ray Spectroscopy (EDS) provides elemental composition information.
Atomic-Scale Structure and Chemistry:
- JEM-ARM300F (GRAND ARM): This aberration-corrected STEM (Scanning Transmission Electron Microscope) achieves resolutions of 58-63 pm, allowing for the direct imaging of atomic structures. With integrated EDS and Electron Energy Loss Spectroscopy (EELS), you can perform detailed chemical analysis of the gate stack and interfaces.
3D Device Tomography:
- JEOL's STEM/EDS tomography capabilities enable the 3D reconstruction of fins and contacts, providing a complete picture of the device's structure.

Application Example: Creating a TEM Specimen of a FinFET

A typical JEOL workflow for creating a TEM specimen of a FinFET might look like this:

SEM Targeting: Use a JEOL SEM to locate the specific fin or feature of interest.
CP or FIB Preparation: Use our Cross Section Polisher™ for a wide, damage-free cross-section, or a FIB-SEM for site-specific milling and lift-out.
Lift-Out and Mounting: The prepared lamella is carefully lifted out and attached to a grid mounted on a double-tilted TEM holder.
HR-STEM Imaging and EELS: The specimen is then transferred to a JEM-ARM300F for high-resolution STEM imaging and EELS mapping of the gate oxide and work-function metals.

Common Pitfalls and How JEOL Mitigates Them

Curtaining and Mechanical Damage: Traditional cross-sectioning methods can introduce artifacts like "curtaining" (vertical lines on the cross-section) and mechanical damage. JEOL's Cross Section Polisher™ (CP) minimizes these effects.
Beam Damage and Contamination: Sensitive materials can be damaged by the electron beam or contaminated by exposure to air. JEOL's air-isolation transfer systems and advanced beam control technologies protect your sample throughout the workflow.

The Takeaway

The performance and reliability of FinFET devices are inseparable from their atomic-scale structure and chemistry. JEOL's comprehensive, end-to-end workflow—from specimen preparation to imaging and analysis—provides the critical insights needed to shorten development cycles and accelerate process learning.

Discuss a FinFET specimen-to-analysis workflow on your device—schedule a session with JEOL applications.

What Is Surface Analysis?

Surface analysis is an analytical technique to elucidate elemental composition, chemical state, and micro structure from material surface layer (several nm to several µｍ). As phenomena such as corrosion, wear, adhesion, and reactions that impact performance and reliability occur primarily on the surface, surface analysis is vital for material evaluation, quality control, and failure analysis.

For analysis, it is necessary to select the most suitable method, according to the sample state (target point, size, material, etc.) and analysis purpose.

What information we want to know? What is the material of the sample?
What is the range of information that we want to know? How deep?
Is it water-soluble? Does it react with solvents? Pre-treatment needed?

It is important to select the analytical method suitable for the purpose.

Types and features of surface analysis instruments

The figure below shows the comparison of typical surface analysis methods from various points of view, such as excitation source, detection signal, quantitativeness, whether the chemical state can be analyzed or not, sensitivity, handling of an insulator, and analysis capability of depth direction. It is important to understand the feature of each method and properly select the analysis method according to the purpose.

Analytical methods	EPMA (WDS)/SXES/EDS	AES	XPS	XRF	SIMS
Excitation source	Electron beam	Electron beam	X-ray	X-ray	Ion
Signal	Characteristic X-ray	Auger electron	Photoelectron	Fluorescence X-ray	Secondary ion
Detectable element	Be ~ (WDS, EDS) Li (SXES, Windowless EDS)	Li ~	Li ~	C ~	H~
Quantitative analysis	○	○	○	○	×
Chemical state	△	○	○	×	Organic compound
Detection depth	Several µm	Several nm	Several nm	Several mm	Several nm
Sensitivity	Several ten ppm （Mass concentration）	Several thousand ppm （Atomic concentration）	Several thousand ppm （Atomic concentration）	Several ten ppm （Mass concentration）	Several ppm （Atomic concentration）
Insulator	○ (Conductive coating)	△	○	○	○
Depth analysis	△	○	○	×	○
Strength	Qualitative analysis Quantitative analysis Wide area ~ micro area analysis	Micro area analysis Chemical bonding state analysis Depth profile analysis	Insulator analysis Chemical bonding state analysis Depth profile analysis	Qualitative analysis Thin film analysis Trace element analysis	Organic substance analysis Trace element analysis
Challenge	Chemical bonding state analysis (Strong at SXES) Organic substance analysis	Wide area analysis Insulator analysis Organic substance analysis	Micro area analysis Trace element analysis	Micro area analysis	Qualitative analysis Quantitative analysis