Minimal-footprint photonic crystal nanolasers for biointegration

TL;DR

The study addresses the bottleneck of bulky, substrate-bound photonic crystal cavities by engineering minimal-footprint, fully detachables for intracellular use. It combines optimized L3 cavities and detached hexagonal PhC nanolaser particles with a complementary twisted-cavity design to achieve lasing in particles as small as $7 μm$ with mode volumes near $0.76 (λ/n)^3$ and Q factors up to about $13{,}000$ in cells. This work enables real-time, nanoscale intracellular sensing with high spatial precision and begins to unlock functionalization avenues for label-free chemical and biological investigations at the single-cell level. The demonstrated platform promises broad impact for biointegration, imaging, and sensing, potentially extending to plasmonic and quantum sensing modalities in biological contexts.

Abstract

Photonic crystals allow unprecedented control over how light is confined, propagates, and interacts with matter. Their development has had a transformative impact on optics and physics, and they remain the central platform for both fundamental discoveries and practical photonic technologies. However, the relatively large footprint and substrate-bound nature of photonic crystal structures have so far strongly limited their use as miniature optical devices or biointegrated sensors. Here, we overcome these limitations by identifying the minimal size of a 2D photonic crystal array needed to achieve lasing and describe the fabrication of substrate-less hexagonal laser particles with an active area as small as 30 μm2. Massively parallel fabrication, robust detachment, and integration of the nanolaser particles into live cells is demonstrated. Crucially, by engineering spatial and spectral mode characteristics, we designed NIR-II probes with mode volumes on the order of tens of attolitres, an order of magnitude smaller than whispering gallery probes of similar dimensions. Such high light localization is comparable in scale to different organelles of eucaryotic cells. In the future, we expect that chemical or plasmonic functionalization of the device will enable label-free sensing of nanoscale intracellular processes, and that it shall serve as a miniature platform to exploit developments in optical and quantum sensing for chemical and biological applications.

Figures (9)

• Figure 1: A small-footprint L3 photonic crystal nanolaser particle of InGaAsP:(a)$|E|^2$ distribution of the fundamental L3 mode of the optimized design on a finite-sized hexagonal particle. The five perturbed holes are highlighted in red. (b) 3D representation of the mode shown in (a), highlighting the strong localization of the field inside the defect area. (c) SEM images of fabricated nanolaser particles. Top: top-view of an optimized L3 cavity, showing the accurate placement of the perturbed holes (scale bar: 100nm). Bottom-left: Detached optimized L3 nanolaser particle resting on the substrate after ultra-sonication (scale bar: 500nm). Bottom-right: Titled view of the defect area (scale bar: 500nm).
• Figure 1: Iterative grid search optimization for the positions of holes adjacent to the defect cavity, to produce a cavity of high quality factor (Q). Instead of starting the optimization from a regular L3 lattice, a five-hole optimization for an L3 cavity performed by Minkov et. al for a silicon resonator around 1.55 was used as the basis Minkov2014. Here, the five adjacent holes were offset from the regular lattice by $\alpha_1 = 0.337a$, $\alpha_2 = 0.270a$, $\alpha_3 = 0.088a$, $\alpha_4 = 0.323a$, and $\alpha_5 = 0.173a$, where $a$ is the lattice spacing. Our simulated device had a lattice spacing of 440 nm, a hole radius of $0.32a$ and a slab thickness of 240 nm, producing a fundamental L3 mode in InGaAsP close to the target wavelength of 1590 nm. In the first iteration, we performed a sweep of the offsets for the first and second holes adjacent to the defect, centred around the basis design. The second iteration used the optimal $\alpha_1$ and $\alpha_2$ found from the first iteration, and swept over offsets for the third and fourth holes. The final iteration again swept over $\alpha_1$ and $\alpha_2$ in a smaller range, producing a final theoretical Q maximum of 2,222,171. The value of $\alpha_5$ was not swept in order to minimize simulation time.
• Figure 2: Determining the minimum array size for the optimized L3 cavity array nanolaser particles.(a) Threshold curves of nanolaser particles showing the emitted power as function of the pump energy for PhC cavities of various sizes. (b) Calculated quality factors corresponding to the threshold curves in (a). (c) Summary of laser threshold energies (triangles) and maximum quality factors (circles) plotted over the array diameter. (d) Lasing spectra at a pump energy of 34.7pJ for nanolaser particles with increasing array diameter. Measured SEM images of three nanolasers are overlaid for size comparison, where the array of adjacent holes (optimized positions highlighted in red) is schematically shown. Colours correspond to the previous panels. For each nanolaser size, the number of air holes adjacent to one side of the L3 defect is also stated. Arrays with less than 4 adjacent holes did not show lasing.
• Figure 2: Laser performance of intracellular lasers with optimized L3 cavities or L3 cavities with a regular hexagonal lattice.(a) Quality factors obtained from measured laser thresholds of nanolasers located inside live fibroblast cells with either the optimized lattice (green) or the regular lattice (purple). The regular design had a lattice spacing of 440 nm and a hexagon side length of 3.5 . The optimized design had a lattice spacing of 420 nm and a side length of 3.6 . Slight adjustments in the size parameters were necessary to match the resonant wavelengths of the two designs. (b) An overview of the maximum quality factors of the measured nanolasers. (c) A comparison of the laser spectrum between the highest quality factor regular lattice nanolaser (purple) and the highest quality factor optimized lattice nanolaser (green).
• Figure 3: L3 nanolaser particles inside live fibroblast cells.(a) Schematic overview of the procedures used to fabricate detached nanolaser particles and their addition to live cell cultures. First, the hole array and trench around each device are defined by e-beam lithography. After dry-etching through all wafer layers, wet-etching the InP substrate layer with hydrochloric acid allows each device to fall onto the etch-stop layer (also InGaAsP). Ultrasonication in PBS detaches the nanolaser particles into solution. Replacing PBS with cell media, the solution is added to cells and incubated for 24 hours to allow for nanolaser internalization. (b) SEM images of nanolasers after wet-etching and brief ultrasonication step (left). Scale bar, 10µm. Longer ultrasonication detaches almost all nanolaser particles (right). Scale bar, 10µm. (c) Normalized lasing spectra from an optimized L3 nanolaser inside a live fibroblast cell, measured at three different pump energies. (d) Brightfield microscopy time-lapse imaging of fibroblast cells and an intracellular nanolaser, covering a period of 4 hours. Scale bar, 100µm. (e) Confocal fluorescence microscopy image of fixed fibroblasts with stained cytosol (purple), nucleus (orange), and a L3 nanolaser (cyan). Scale bar, 10µm.