Micro-LED displays have rapidly become one of the most compelling technologies in advanced display engineering. By using inorganic light-emitting diodes shrunk to microscopic dimensions, they deliver exceptional brightness, fast response times, strong contrast, and remarkably low power consumption. These properties make them especially attractive for near-eye applications such as AR and VR headsets, where pixels must be both extremely small and extremely efficient. Yet building a full-color micro-LED display remains one of the field’s most stubborn engineering challenges, because red, green, and blue emitters must all be fabricated and aligned with near-perfect registration at the micrometer scale and below.

One of the most practical strategies for achieving full color is quantum-dot color conversion. In this scheme, the display is built from efficient blue micro-LEDs, and selected pixels are overlaid with films of red- or green-emitting quantum dots—nanoscale semiconductor crystals whose emission wavelength is set by their size. When blue light from the underlying LED excites the dots, they re-emit it as red or green, while the untouched blue pixels provide the third primary color. The approach elegantly sidesteps the difficulty of making efficient red micro-LEDs directly, since red indium gallium nitride emitters suffer from sharply declining efficiency as their size shrinks.

The bottleneck, however, lies in patterning. Different quantum dots must be placed into precisely defined, uniform, and closely spaced pixel wells without contaminating neighboring pixels, spreading across the surface, or degrading the delicate optical properties of the nanocrystals themselves. Conventional patterning techniques—photolithography, inkjet printing, transfer stamping—each bring trade-offs involving multi-step processing, limited resolution, material bleed, or chemical and thermal damage to the quantum dots.

The Beijing Institute of Technology team, led by Professor Xueqiang Zhang with PhD student Muhammad Farhan as first author, addressed this problem with a spatially shaped femtosecond laser combined with a selective filling process. Their workflow begins with a glass substrate coated with a thin layer of SU-8, a common epoxy-based photoresist polymer. Using a femtosecond laser, the researchers drill ordered arrays of microholes directly into the polymer, creating an empty pixel mold before any quantum dots are ever introduced.

The key innovation lies in how the laser beam is shaped. A conventional Gaussian beam concentrates its intensity in a smooth, bell-shaped profile that tends to produce holes with ragged entrances, rough walls, and a characteristic taper that widens toward the surface. Moreover, high intensities deep in the focus can damage the underlying substrate. The team instead employed a Bessel beam, a non-diffracting beam profile with an elongated axial focus that distributes the laser energy differently. Compared with Gaussian drilling, the shaped Bessel beam produced microholes with cleaner entrances, smoother sidewalls, reduced taper, and markedly less damage to the glass below. By tuning laser parameters such as pulse energy and focusing conditions, the researchers could sweep the hole diameter across a broad range, ultimately demonstrating well-formed nanoholes as small as 100 nanometers—a scale at which the feature is smaller than most wavelengths of visible light.

Once the mold was in place, the delicate luminescent material entered the picture. Red or green cadmium selenide (CdSe) quantum dots were deposited into selected holes and cured in place, after which excess material was removed from the surface, leaving clean, isolated pixel wells. Because the femtosecond laser never touches the quantum dots directly—the light-emitting nanocrystals are introduced only after the drilling is complete—the method avoids the photochemical damage and thermal stress that laser-based patterning often inflicts on quantum materials. This separation of steps is central to the technique’s appeal: the laser provides lithographic precision while the quantum dots retain their pristine optical quality.

The team demonstrated the versatility of the approach by fabricating well-defined monochromatic red and green pixel arrays, with individual luminescent pixels reaching diameters of approximately 700 nanometers. By performing sequential rounds of drilling and filling—first patterning one set of wells and populating them with one color of dots, then drilling a second aligned array and filling it with another—the researchers also produced red-and-green dual-color arrays with crisp, well-resolved boundaries between neighboring pixels of different colors. Such spatial registration is precisely what a full-color display demands.

The optical performance of the resulting color-conversion layers was equally impressive. The patterned films exhibited narrow-band red and green fluorescence, indicating high color purity—a critical figure of merit for wide-gamut displays. Pixel-to-pixel emission was highly uniform, with the red and green arrays achieving luminous uniformities of approximately 90 percent and 97 percent, respectively. Taken together, the color performance of the fabricated layers corresponded to a color gamut covering 111 percent of the NTSC standard, exceeding even the benchmark set for high-end display color reproduction.

What makes the work particularly notable is the combination of attributes that rarely coexist in a single patterning method: it is maskless, requiring no expensive photomasks and permitting rapid design iteration; it is solvent-free in the patterning step, simplifying processing; and it is fundamentally quantum-dot-compatible, since the light-emitting material is shielded from all direct laser exposure. The resolution—sub-micrometer pixels and 100-nanometer holes—far exceeds what most printing-based approaches can achieve, while the smooth, low-damage hole walls produced by the Bessel beam ensure that filled dots emit efficiently and uniformly.

The researchers acknowledge that the path from laboratory demonstration to manufacturing will require further development. Scaling the approach to large display areas will demand parallel laser processing, since sequentially drilling billions of pixel wells with a single beam is impractical for mass production. Improvements in quantum-dot filling efficiency and in integration with actual micro-LED driver backplanes will also be essential before the technique can populate real devices. Yet the demonstration of reliable, uniform, sub-micrometer luminescent pixels with excellent color purity establishes a flexible and promising foundation for that scale-up effort.

If those engineering challenges can be met, the implications extend beyond AR and VR headsets. High-pixel-density quantum-dot color-conversion layers could benefit microdisplays for projectors, heads-up displays, ultra-dense sensor arrays, and any application where light of different colors must be generated in pixels smaller than the eye can resolve. For now, the Beijing team’s spatially shaped femtosecond laser offers a vivid illustration of how advances in light-shaping and precision manufacturing continue to push the physical limits of what a display pixel can be.