Access Official Technical PDF

Share Summary:LinkedInEmail

Nanowire substrate reuse could help address one of the cost constraints associated with III-V semiconductor nanowire production: the reliance on expensive single-crystal wafers that are commonly used for a single growth cycle. A demonstrated approach uses an aluminum phosphide (AlP) segment beneath gallium phosphide (GaP) nanowires. The AlP acts as a sacrificial layer, enabling nanowire removal and selective cleaning of the growth surface before another cycle.

The method was investigated using vapor-liquid-solid (VLS) growth by metalorganic vapor-phase epitaxy (MOVPE). Results included successful axial AlP/GaP nanowire growth and two substrate reuse cycles. They show that a sacrificial layer can support regrowth, while also underscoring the importance of growth, harvesting, etching and seed-particle preparation as parts of one integrated process.

Why nanowire substrate reuse matters

Semiconductor nanowires have potential in applications including photonics, sensors, energy technologies and transistors. MOVPE can produce nanowires with controlled structural, electrical and optical properties, but the process uses costly III-V substrates and is typically batch-based. When a wafer cannot be used again, its cost contributes to the expense of each production cycle.

Reusing the same substrate could reduce the amount of new wafer material needed. The challenge is to remove the grown nanowires without leaving a surface that prevents reliable regrowth. Simply detaching nanowires may leave stubs or other residues, while removing substrate material during cleanup can limit how many times a wafer remains usable.

How the AlP sacrificial layer enables reuse

The approach places an AlP segment at the base of each GaP nanowire. The two materials form an axial heterostructure: the AlP section is grown first on the GaP wafer, followed by the GaP nanowire segment. Gold seed particles define the nanowire growth sites, and a silicon nitride (SiN) mask helps position those particles on the substrate.

After growth, the GaP nanowires are mechanically harvested. This leaves AlP residues at the growth sites. A selective wet chemical etch is then used to remove the AlP, with the goal of clearing the holes in the SiN mask while preserving the GaP wafer and mask. The substrate can then be prepared with seed particles for another growth cycle.

In practical terms, the reuse sequence consists of:

  • Patterning a SiN mask and defining gold seed-particle sites.
  • Growing axial AlP/GaP nanowires by MOVPE.
  • Mechanically removing the nanowires for downstream use.
  • Etching away AlP residues and preparing the substrate for regrowth.

Each step affects the next. For example, residue removal must be effective enough to restore the growth sites, but the etch must also avoid damaging the underlying wafer or mask.

Growth and reuse results

Growth optimization produced maximum reported yields of 96% for GaP nanowires, 96% for AlP nanowires and 97% for axial AlP/GaP nanowires. In the reuse experiments, regrowth of axial AlP/GaP nanowires achieved an 80% yield on 1 × 1 cm² substrates. The overall yield was 82% for the first reuse cycle and 75% for the second.

The substrates remained intact after the second regrowth, indicating that further reuse may be possible. However, these results represent two cycles, not a demonstration of indefinite reuse. Yield also varied between cycles, so the results should be understood as evidence of feasibility rather than a guarantee of unchanged performance over repeated production.

What a conventional cleanup comparison shows

A reference experiment explored reuse without a sacrificial layer. A GaP wafer that had previously supported GaP nanowire growth was wet-etched to planarize the surface before processing and regrowth. GaP nanowire regrowth was possible, with a reported yield of 50%.

This comparison illustrates a central process distinction. Without a separate sacrificial segment, surface preparation relies on removing material from a wafer that may contain nanowire stubs. The AlP-based method instead targets the AlP residue selectively, with the aim of restoring the growth sites without substantial removal of the GaP substrate. The reported results do not establish that every wafer or process will behave identically, but they show why selective removal is important to the reuse strategy.

Process considerations for engineering teams

For teams assessing this approach, the key question is not only whether a nanowire can be detached, but whether the complete substrate cycle can be repeated consistently. Several process controls are central:

  • Growth conditions: The AlP base and GaP nanowire must both grow successfully in the axial structure.
  • Harvesting: Mechanical removal must leave the substrate in a condition that can be cleaned and processed again.
  • Etch selectivity: The AlP residue must be removed while protecting the GaP surface and SiN mask.
  • Seed definition: Gold seed particles need to be re-established at suitable growth sites for regrowth.
  • Yield tracking: Initial growth yield and yield after reuse are distinct measures and should be evaluated separately.

The work also highlights that wafer reuse is a materials and integration problem, not just a growth problem. Mask condition, residue removal and seed-particle redeposition all influence whether a substrate can return to the growth process.

Practical implications

An AlP sacrificial layer provides a demonstrated route to harvesting GaP nanowires and returning the underlying wafer to regrowth. The results—two reuse cycles, intact substrates after the second cycle and yields of 82% and 75% for the respective cycles—support further investigation while leaving questions about long-term repeatability and process consistency open.

For industrial and integration teams, the practical takeaway is to evaluate the full cycle as a coordinated process: grow the heterostructure, harvest the nanowires, selectively remove the sacrificial material and prepare the growth sites again. If those operations can be controlled, substrate reuse may offer a way to reduce dependence on single-use III-V wafers without treating the wafer as a consumable after every nanowire run.

Access Official Technical PDF

About The Author