Writing in Light: Science & Applications, Peng and colleagues report a device they describe as a robust single-mode laser based on merging bound states in the continuum, or BICs. Their platform is a photonic-crystal slab laser built on InGaAsP multiple quantum wells. By tuning the diameter of the air holes etched into the slab, the researchers move accidental BICs in momentum space and merge them with a symmetry-protected BIC at the Gamma point, the center of the Brillouin zone. In an ideal, infinite photonic crystal, this merging broadens the high-Q region around Gamma: instead of a narrow, isolated high-Q point, radiative loss remains suppressed over a larger domain of in-plane wave vector. The concept builds on the topological picture of BICs as polarization singularities carrying conserved topological charges, and on the earlier experimental demonstration that merging multiple BICs can create ultrahigh-Q guided resonances that are robust against out-of-plane scattering.

At first glance, one might expect the exact merging point to be the optimum laser design, because it delivers the broadest radiative protection. Peng and colleagues show that this intuition is incomplete, and their correction is the most consequential finding of the study. In a finite photonic crystal, the continuous BIC band of the ideal infinite slab is quantized into discrete, cavity-like modes. The desired fundamental state and the nearby higher-order states are therefore not independent of the BIC landscape; they both sample it. If the high-Q region becomes too broad, it can protect the competing modes just as effectively as the target mode. The result is a lower absolute threshold, but a weaker single-mode hierarchy, because the very mechanism that shields the wanted mode also shields its rivals.

The key message of the paper is that the most robust single-mode laser is obtained not exactly at the merging condition, but slightly before it. In this pre-merging regime, the fundamental s-like BIC-derived mode is already strongly protected, while the nearest p-like competing mode still experiences substantially larger loss. The authors calculate that in the pre-merging device, the threshold gain of the p-like competitor is more than three times larger than that of the s-like lasing mode. At exact merging, this contrast collapses to only about 1.3. Experimentally, the difference is decisive: the pre-merging device remains strictly single-mode up to 80 times threshold, whereas at the merging condition a higher-order BIC-derived mode begins to lase at high pump power. The pre-merging point is, in effect, a sweet spot where protection is strong enough for the target mode but no longer generous enough for its competitors.

This reframes the role of BIC merging in active devices. Merging BICs were originally attractive because they suppress radiative leakage and make high-Q resonances more robust against fabrication disorder and finite angular spread of the pump. Super-BIC lasers have already exploited this principle to reduce threshold by merging symmetry-protected and accidental BICs in momentum space. What Peng and colleagues add is an explicitly active-laser criterion: one should not maximize Q blindly, but should instead maximize the useful threshold contrast between the desired mode and the nearest competing modes. The relevant figure of merit is therefore not simply the quality factor itself, but something closer to the ratio of the modal gains of the target and competing modes, or more generally the separation between the first and second lasing thresholds under realistic gain saturation.

The distinction matters most for miniaturized BIC lasers. Ideal BICs are Bloch modes of extended periodic structures, and their high-Q character is usually compromised when the device is made small. Reducing the lateral size increases edge leakage, broadens the momentum distribution of the confined field, and weakens the infinite-lattice protection that makes BICs attractive in the first place. Early BIC lasing experiments demonstrated that lasing action could persist in finite arrays, even down to a few periods of the crystal. The broader mini-BIC problem, however, is more subtle: how can one retain BIC-like vertical confinement while also providing in-plane confinement in a genuinely small footprint, without sacrificing the very quality factors that motivated the design?

Several research groups have addressed this question by combining BIC physics with additional lateral confinement mechanisms. One team demonstrated miniaturized BIC cavities with ultrahigh quality factors by combining lateral photonic-bandgap confinement with vertical BIC protection, reaching very high Q in small modal volumes. Others then used mini-BIC cavities with quantum-dot gain media to demonstrate low-threshold single-mode nanolasers, and subsequently achieved ultra-low-threshold continuous-wave operation with quantum-dot mini-BIC lasers. A further approach achieved confinement through topological band inversion, producing electrically pumped compact bulk lasers driven by band-inverted bound states in the continuum. These works established that mini-BICs can overcome the apparent incompatibility between BIC physics and small mode volume, but they often rely on a designed cavity region embedded within a larger photonic environment rather than on the patterned region alone.

Peng and colleagues attack the problem from a different angle. In their device, the entire patterned photonic-crystal region spans only five by five periods, an area of roughly 15 square micrometers. At this scale, one cannot simply assume that the structure behaves as a truncated version of an infinite BIC lattice; edge leakage becomes a dominant part of the cavity physics. The authors therefore engineer the boundary holes, reducing their diameter relative to the central holes, to suppress leakage and recover BIC-derived lasing in an extremely small patterned region. The significance is not merely that a small cavity tends to have fewer modes. Reducing the area generally increases mode spacing and can make single-mode operation easier in a trivial spectral sense. The real challenge is different: a very small BIC laser may lose the high-Q advantage that made the BIC useful in the first place. The new work shows that merging-BIC physics, finite-size quantization and edge engineering can be combined so that compactness does not simply destroy BIC protection.

The findings also connect to a parallel development in passive BIC photonics. Recently, the merging of accidental BICs with net-zero topological charge was shown to reshape the polarization texture of the emitted field, stretching the range of circularly polarized emission over a sizeable momentum-space domain. Taken side by side with the new laser result, this suggests that BIC engineering may eventually move beyond selecting a single frequency and spatial mode, toward selecting a single handedness or spin channel. That would open a possible conceptual route toward chiral or spin-selective BIC lasers, in which the same topological machinery that suppresses radiation loss also dictates the polarization state of the output beam.

The broader implication is that BIC lasers are entering a regime where topology, finite-size physics and gain competition must be designed together. Passive BIC photonics asks how radiation can be canceled. Active BIC photonics asks a harder question: after radiation is canceled for one state, which other states are also helped, and will any of them lase? Peng and colleagues answer by identifying a pre-merging sweet spot, demonstrating that the best laser is not located at the point of maximum radiative protection, but at the point where radiative protection is most selective. That principle, established here in a photonic-crystal slab small enough to fit comfortably within a few wavelengths of light, may become a standard design rule for compact, stable and eventually electrically pumped BIC lasers, and a reminder that in laser engineering, as in many fields, the optimum rarely coincides with the maximum.

Subject of Research: Modal selectivity in compact merging-bound-state-in-the-continuum photonic-crystal lasers

Article Title: Single-mode lasers before the merger

Article References: Zito, G. (2026). Single-mode lasers before the merger. Light: Science & Applications, 15(1), Article 372. https://doi.org/10.1038/s41377-026-02467-3

Image Credits: AI Generated

DOI: 10.1038/s41377-026-02467-3

Keywords: bound states in the continuum, photonic crystals, semiconductor lasers, single-mode lasing, quality factor, modal selectivity, InGaAsP quantum wells, miniaturized lasers, topological photonics, threshold discrimination, gain competition, nanophotonics