🔬 optics

Colloidal Nanocrystals Regrowth-Assisted Synthesis of Perovskite Microwire Lasers for Integrated Optoelectronics

This paper reports a straightforward colloidal synthesis method using diphenyl ether to produce high-quality CsPbBr3 microwire lasers that, after ion exchange for spectral tunability, are successfully integrated into a lossless nanowaveguide device demonstrating nonlinear optoelectronic responses suitable for on-chip neuromorphic computing.

Original authors: Elizaveta V. Sapozhnikova, Ivan A. Matchenya, Dmitry A. Tatarinov, Grigorii A. Verkhogliadov, Dmitry A. Semyonov, Maria A. Kirsanova, Natalia K. Kuzmenko, Julia S. Mironova, Arina O. Kalganova, Valeri

Published 2026-03-26

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Elizaveta V. Sapozhnikova, Ivan A. Matchenya, Dmitry A. Tatarinov, Grigorii A. Verkhogliadov, Dmitry A. Semyonov, Maria A. Kirsanova, Natalia K. Kuzmenko, Julia S. Mironova, Arina O. Kalganova, Valeriya M. Levkovskaya, Stepan A. Baryshev, Yuxi Tian, Anatoly P. Pushkarev

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Building a Tiny, Tunable Light Show

Imagine you are an architect trying to build a miniature city of light. You need tiny, perfect crystal "buildings" (microwires) that can act as both a laser (a light source) and a detector (a light sensor). The problem is, making these crystals usually results in a messy pile of rubble or misshapen blobs.

This paper describes a new, clever recipe for growing these tiny crystal buildings perfectly, tuning their color like a radio, and then wiring them together to create a super-fast, brain-like computer chip.

1. The Secret Ingredient: The "Glue" and the "Sacrifice"

The Problem: Usually, when scientists try to grow these crystals, they use a lot of "ligands" (think of these as sticky hands that hold the crystals apart). If you have too many sticky hands, the crystals stay small and round. If you have too few, they clump together into a messy ball.

The Solution: The researchers changed the "soup" (solvent) they used. Instead of the usual oil, they used Diphenyl Ether (DPE).

The Analogy: Imagine DPE is a very strong magnet for the lead atoms in the mix. Because the solvent holds the lead so tightly, the scientists didn't need to add as many "sticky hands" (ligands).
The Result: This allowed them to control the growth perfectly.

The "Sacrificial" Step:
When they mixed the ingredients, they didn't get the final product immediately. They first made a "sacrificial" type of crystal (Cs4PbBr6).

The Analogy: Think of this like a sugar cube dissolving in tea. The sugar cube isn't the tea you want to drink, but as it dissolves, it slowly releases sugar molecules that help the tea become sweet.
What happened: These "sugar cube" crystals slowly dissolved, releasing building blocks that slowly grew into the perfect, long, thin wires (microwires) needed for lasers.

2. From Tiny Dots to Long Wires: The "Velcro" Effect

Once the building blocks were ready, they started growing.

Nanocrystals: Tiny dots.
Nanowires: The dots lined up like beads on a string.
Microwires: The strings grew longer and thicker.

The Magic of Stirring:
The researchers found that stirring the pot was crucial.

The Analogy: Imagine a crowd of people trying to form a line. If they just stand there, they might bunch up randomly. But if you gently push them (stirring), they align themselves.
The "Chipping": The stirring also helped snap off the rough, uneven ends of the wires. This is critical because a laser needs two perfectly flat mirrors at the ends to bounce light back and forth. The stirring naturally "chipped" the ends to make them flat, creating a perfect Fabry-Pérot cavity (a light trap).

3. Tuning the Color: The "Radio Dial"

The team wanted to change the color of the laser light. They used a process called Ion Exchange.

The Analogy: Imagine the crystal wire is a sponge soaked in red dye (Bromine). They dipped it into a solution containing a "blue dye" (Chlorine) and a special helper (Yttrium).
The Process: The helper (Yttrium) acted like a gentle guide, helping the blue dye slowly swap places with the red dye inside the crystal.
The Result: They could tune the laser from green (525 nm) to blue-green (485 nm) just by changing how much "blue dye" they added. It's like turning a radio dial to find a different station.

4. The Grand Finale: The "Neuromorphic" Device

Finally, they built a tiny machine to prove these wires are useful.

The Setup: They took a Laser Wire (the light source), placed it on top of a Waveguide (a glass fiber that carries light), and connected it to a Photodetector Wire (a sensor that turns light into electricity).
The Analogy: Think of it as a relay race.
1. The Laser Wire shouts a message (light).
2. The Waveguide is the track the message runs along.
3. The Photodetector catches the message and turns it into a signal.

The "Brain" Connection:
The most exciting part is how this device behaves. When they hit it with light and electricity at the same time, it didn't just react linearly (1 + 1 = 2). It reacted non-linearly.

The Analogy: A normal light switch is "on" or "off." This device is more like a human brain neuron. It has a "memory." If you send a weak signal, it might ignore it. But if the signal gets strong enough (crossing a threshold), it suddenly fires a big response and remembers that it was triggered for a little while after the signal stops.
Why it matters: This "memory" and "threshold" behavior is exactly what computers need to mimic the human brain (neuromorphic computing). It could lead to chips that are faster and use less energy than today's computers.

Summary

In short, these scientists:

Found a new way to grow perfect crystal wires by using a special solvent and a "sacrificial" intermediate step.
Learned how to tune their color like a radio dial.
Built a tiny, integrated circuit that acts like a brain cell, capable of processing light and electricity in a way that mimics human memory.

This is a major step toward building super-fast, light-based computers that fit on a single chip.

1. Problem Statement

While colloidal halide perovskite nanocrystals (NCs) are well-established for light-emitting applications, synthesizing high-quality, one-dimensional microwires (MWs) suitable for lasing and integrated optoelectronics remains challenging. Existing methods often suffer from:

Polycrystalline structures and irregular cavity shapes.
Coalescence of particles in solution, leading to agglomerated clusters rather than discrete wires.
Difficulty in controlling the transition from quantum-confined nanowires (NWs) to bulk microwires without degrading optical quality.
Lack of precise control over the emission wavelength (bandgap tuning) in the final microwire form.

These issues hinder the large-scale, low-cost production of photonic integrated circuits and neuromorphic computing devices based on perovskites.

2. Methodology

The authors developed a novel colloidal synthesis strategy utilizing a regrowth-assisted mechanism to transform nanocrystals into microwires.

Solvent Engineering: Replaced the standard 1-octadecene (ODE) solvent with diphenyl ether (DPE). DPE coordinates strongly to $Pb^{2+}$ ions, allowing for the dissolution of $PbBr_2$ with a significantly reduced excess of organic ligands (oleic acid/oleylamine). This reduction prevents the inhibition of NC regrowth.
Two-Stage Synthesis Process:
1. Precursor Formation (140°C): Injection of cesium oleate (CsOA) into the $PbBr_2$ /ligand/DPE solution yields a mixture of small orthorhombic $CsPbBr_3$ NCs and large, sacrificial rhombohedral $Cs_4PbBr_6$ NCs.
2. Regrowth and Transformation (140°C $\to$ 180°C):
  - Stage 1 (140°C, 2h): The sacrificial $Cs_4PbBr_6$ phase slowly dissociates, releasing perovskite precursors. This allows for the gradual growth of $CsPbBr_3$ NCs into nanowires (NWs) via an oriented-attachment mechanism.
  - Stage 2 (180°C, 1h): Further heating promotes the assembly of NWs into needle-like structures, which align and fuse into microwires. Mechanical stirring is critical here; it applies shear forces to break long arrays into needles and induces collisions that "chip" the uneven ends of the wires, creating flat, reflective facets essential for Fabry-Pérot (FP) laser cavities.
Post-Synthetic Tuning: To tune the emission wavelength, the authors employed ion exchange using $YCl_3$ in a DPE:hexane solution. The slow diffusion of halide ions allowed for the controlled substitution of Br with Cl, shifting the bandgap without destroying the microwire morphology.
Device Integration: A hybrid device was assembled using a dry-transfer technique:
- Source: A $CsPb(Cl,Br)_3$ microwire laser.
- Waveguide: An $\alpha-MoO_3$ nanowaveguide (lossless, transparent in the emission range).
- Detector: A $CsPbBr_3$ microwire photodetector on gold electrodes.

3. Key Contributions

Mechanistic Insight: The study provides a comprehensive, step-by-step reconstruction of the colloidal evolution from NCs to NWs to MWs, identifying the role of the sacrificial $Cs_4PbBr_6$ phase and the critical importance of stirring in forming discrete, high-quality laser cavities.
High-Quality MW Lasers: The synthesis produces $CsPbBr_3$ microwires (5–20 µm length) with excellent crystallinity, smooth end facets, and sub-wavelength cross-sections, enabling low-threshold, high-Q lasing.
Spectral Tunability: Demonstrated the ability to tune the lasing emission from 525 nm to 485 nm via $YCl_3$ -mediated ion exchange, creating mixed-halide $CsPb(Cl,Br)_3$ lasers.
Integrated Neuromorphic Device: Successfully fabricated a compact, all-perovskite integrated circuit where a microlaser couples to a nanowaveguide and excites a photodetector, exhibiting nonlinear optoelectronic responses suitable for neuromorphic computing.

4. Key Results

Structural Characterization:
- XRD and TEM confirmed the transformation from $Cs_4PbBr_6$ to highly crystalline orthorhombic $CsPbBr_3$ .
- HAADF-STEM revealed that MWs are composed of aligned NWs (approx. 30 nm width) with lattice-matched boundaries and minimal defects.
- End facets are naturally chipped by stirring, creating effective FP cavities.
Optical Performance:
- Lasing: Pristine $CsPbBr_3$ MWs exhibit low-threshold lasing ( $F_{th} \approx 27.5 \, \mu J/cm^2$ ) with a high Q-factor (~3030).
- Tuning: $CsPb(Cl,Br)_3$ MWs showed tunable lasing between 485–540 nm. While Cl-doping reduced PLQY and increased the lasing threshold (up to ~70 $\mu J/cm^2$ ), high-Q lasing was maintained.
- Exciton-Polaritons: Spectral modulation indicated the presence of exciton-polaritons in the microcavity.
Device Functionality:
- The integrated device (Laser $\to$ $MoO_3$ Waveguide $\to$ Photodetector) successfully transmitted coherent light over ~84 µm.
- Nonlinear Response: Under hybrid optical-electrical stimulation, the device exhibited a "pre-pulse" current and a step-like photocurrent jump upon lasing threshold crossing.
- Neuromorphic Behavior: The charge vs. pump power dependence showed nonlinearity orders of 0.9, 1.5, and 3.2 for biases of 1V, 2V, and 4V, respectively. This behavior is attributed to trap-filling dynamics and Schottky barrier modulation, mimicking synaptic plasticity (volatile and non-volatile memory).

5. Significance

This work represents a significant leap forward in integrated perovskite optoelectronics.

Scalability: The colloidal synthesis method is solution-processable and avoids complex lithography for the active material, offering a pathway to low-cost, large-scale production of photonic components.
Material Quality: The "regrowth-assisted" approach solves the historical problem of obtaining discrete, high-quality microwire lasers from colloidal solutions.
Functional Integration: By demonstrating a fully functional, on-chip device where a laser drives a photodetector via a waveguide, the authors bridge the gap between material synthesis and system-level application.
Neuromorphic Computing: The observed nonlinear optoelectronic response suggests these devices can serve as hardware components for neuromorphic computing, potentially replacing traditional electronic neurons with faster, more energy-efficient photonic counterparts.

In conclusion, the paper establishes a robust framework for synthesizing tunable perovskite microwire lasers and integrating them into complex optoelectronic circuits, paving the way for advanced on-chip photonic and neuromorphic technologies.