Wien Filter Correction of Chromatic Deflection Aberration for Large-Field High-Resolution LVSEM: A Practical and Cost-Effective Implementation

Low-voltage scanning electron microscopy (LVSEM) has emerged as an essential technique for imaging fragile nanoscale materials such as photoresists, soft polymers, and 2D semiconductors that are highly sensitive to beam damage. By operating at landing energies below 1 keV, LVSEM minimizes sample charging and structural damage, while still delivering high-resolution surface information. However, a persistent challenge has been extending high-resolution performance across wide fields of view (FOV) without sacrificing speed or stability. To avoid slow and often mechanically noisy sample stage movement, many SEM systems now rely on beam-image shift technique, where the electron beam is deflected electronically to scan different regions of the sample. This method is efficient and eliminates mechanical lag, but it comes with a trade-off: shifting the beam introduces chromatic deflection aberration. This distortion arises because electrons within the beam don’t all have exactly the same energy, and as a result, they respond slightly differently to deflection fields. The wider the energy spread or the greater the deflection, the worse the blurring becomes—especially at the outer regions of the scan area. Several strategies have been developed to combat this issue, but each has its downsides. Cold field emission guns, for example, reduce energy spread but are prone to instability during long sessions. Monochromators can narrow the beam energy distribution but do so at the cost of reducing current, which slows imaging. Meanwhile, high-end solutions like aberration correctors or moving objective lenses offer excellent correction but are expensive, complex, and not easily implemented in most SEM setups. This leaves a gap in the field: a need for a simpler, more accessible way to tackle chromatic aberration without overhauling the instrument.

Recognizing this, new research paper published in Journal of Vacuum Science & Technology B and led by Shun Kizawa, D. Bizen, K. Suzuki, S. Mizutani, R. Watanabe, Y. Kasai, Y. Mizuhara from Hitachi Ltd in Japan, developed a new approach centered around integrating a Wien filter directly into the SEM column. Unlike conventional corrective optics, the Wien filter uses crossed electric and magnetic fields to selectively counteract the chromatic deflection aberration introduced during beam shifts.
To test its performance, the team developed a customized LVSEM platform equipped with a Schottky emission source, a precision beam deflector, and their newly implemented Wien filter module. They conducted systematic imaging tests across a 28 × 28 µm² field of view using black silicon as a test sample. This surface, known for its finely textured features, served as a reliable benchmark for resolution degradation.
Without correction, the resolution degraded by as much as 16% at the edges compared to the center—a level of blur that can significantly impact measurements. Once the Wien filter was engaged and calibrated according to their model, that degradation dropped to around 8%.  To further assess the correction’s effectiveness, the team analyzed resolution anisotropy by plotting degradation as a function of edge orientation. Prior to correction, the results revealed elliptical deformation, consistent with directional chromatic aberration. After correction, this pattern disappeared, replaced by a weaker, asymmetrical distortion more characteristic of coma aberration. While the Wien filter wasn’t designed to correct for coma, the remaining distortion was minor, especially in comparison to the uncorrected chromatic effects. An important part of their study also focused on response time—specifically, how quickly the Wien filter could respond relative to the beam-image shift itself. This matters because any lag in correction could introduce its own artifacts, undermining the benefit. Using a time-resolved scan over a line-and-space pattern, the researchers measured the time it took for the image to stabilize after initiating a shift. Remarkably, the Wien filter reached its steady state 60% faster than the beam deflector.

To evaluate how well their chromatic aberration correction method would perform in real-world measurement tasks, the authors tested it on a patterned semiconductor sample—a dot array formed on a silicon wafer. These kinds of structures are routinely used in semiconductor metrology because they offer well-defined geometries that are sensitive to changes in imaging resolution. Using their modified LVSEM system, the researchers measured the critical dimensions (CD) of the dots at multiple beam-image shift positions across the field of view. This allowed them to assess how uniform the measurements remained when the beam was deflected to various off-axis regions.
Uncorrected, the data revealed noticeable inconsistencies. The variation in CD measurements, quantified using the standard three-sigma (3σ) metric, reached 0.67 nm across the scanned area—an amount that could easily affect yield in semiconductor manufacturing where dimensional tolerances are extremely tight. Once the Wien filter correction was applied, this variability dropped to 0.40 nm, representing an improvement of about 40%. That reduction is significant, especially considering how subtle these discrepancies can be and how difficult they are to address without specialized correction hardware.
Moreover, the Hitchai team looked at how the correction held up across a range of edge-detection settings—from low to high thresholds. At the commonly used 50% threshold, which corresponds to the midpoint in grayscale intensity between a feature and its background, the uniformity still improved by about 20%. What sets this study apart is not just the quality of the results but the practicality of the solution. The authors tackled a persistent limitation in LVSEM—maintaining high resolution across wide scanning areas—without resorting to expensive or cumbersome modifications. Instead of relying on monochromators (which reduce beam current), aberration correctors (which are difficult to calibrate), or custom objective lenses (which are expensive to manufacture), they implemented a compact Wien filter that integrates easily into conventional SEM systems. This design choice keeps the hardware manageable while offering dynamic compensation for chromatic deflection aberrations.

In conclusion, the implications of the research work of Shun Kizawa  and colleagues for metrology are substantial because as the semiconductor industry moves toward ever-smaller nodes and tighter process controls, the demand for fast, reliable imaging tools is only increasing. This correction method offers a direct path to improved measurement accuracy, especially in critical dimension metrology where nanometer-level precision is non-negotiable. In processes like EUV lithography, where stochastic variability already pushes measurement systems to their limits, the ability to reduce noise and maintain uniformity could have a measurable impact on yield and device performance. Another key advantage is the filter’s fast response time. In high-throughput applications, the imaging system must adapt almost instantaneously as the beam shifts from one region to another. The Wien filter’s quick stabilization—faster than the beam deflection itself—makes it ideal for real-time corrections. This avoids the need for added delays or post-acquisition compensation, streamlining the entire imaging workflow.

Additionally, the new correction system could benefit materials science, nanotechnology, and even biological imaging, where large-area scans are often needed but beam-sensitive samples limit the use of higher landing energies. By operating effectively at low voltages, the method helps preserve fine surface details and reduces issues like sample charging or damage—common challenges when imaging polymers, biomaterials, or 2D structures. Perhaps most exciting is how this work lays the foundation for future enhancements. While the Wien filter effectively suppresses chromatic deflection aberration, some residual coma remains, as shown in the direction-dependent resolution plots. This opens the door to incorporating additional correction optics, such as sextupole lenses, to tackle the next layer of distortion. The broader principle—real-time, energy-sensitive correction—could inspire similar innovations in other particle-based imaging systems, from focused ion beams to X-ray optics.