Precise Control of Non‑Close‑Packed Polystyrene Nanoparticle Arrays via Ion Beam Etching

Background

Polystyrene nanospheres have emerged as versatile building blocks for nanostructure fabrication, enabling the creation of ordered nanowire, nanopillar, nanohole, nanodot arrays, core/shell composites, nanomeshes, and magnetic quantum dots [1–13]. Nanosphere lithography is particularly attractive due to its simplicity and low cost. Conventionally, a monolayer of PS spheres self‑assembles into a hexagonal close‑packed (HCP) lattice on a planar substrate via spin‑coating or Langmuir–Blodgett deposition [14,15]. Reducing the sphere diameter while preserving the lattice position transforms the HCP arrangement into a non‑close‑packed (NCP) array, which can subsequently be used as a mask for etching ordered nanostructures [1–12]. The achievable feature size and inter‑particle spacing depend critically on the initial sphere diameter and the etching parameters.

Traditional etching methods—reactive ion etching (RIE) with O₂ plasma and plasma etching (PE) with Ar plasma—operate at high rates (≈40–90 nm min⁻¹ for RIE, ≈180 nm min⁻¹ for PE) and are inherently anisotropic, leading to non‑spherical particle morphologies and poor control below 300 nm [16–21]. Recent isotropic techniques, such as inductively coupled plasma etching (ICPE), achieve slower rates (~8 nm min⁻¹) by independently tuning plasma density and bias voltage, enabling sub‑50 nm control while maintaining spherical shapes [22,23].

Ion beam etching (IBE) offers distinct advantages: the ion source and acceleration stages are decoupled from the substrate, minimizing lateral plasma bombardment and enabling independent control of ion current density and energy. To date, NCP PS arrays produced by IBE have not been reported. This study addresses that gap by demonstrating precise size control of PS nanospheres via IBE and subsequent fabrication of ordered Si nanopillars.

Methods

Commercial 100 nm PS nanospheres (Alfa Aesar, 2.5 wt % suspension) were deposited onto RCA‑cleaned p‑type Si(100) wafers using the Langmuir–Blodgett technique to form a monolayer HCP lattice. After drying, the samples were loaded into a vacuum chamber (pressure < 6.0 × 10⁻⁴ Pa). An Ar⁺ ion beam (Kaufman source) was generated at a chamber pressure of 2.0 × 10⁻² Pa and directed normal to the substrate. Various etching conditions (time, beam current, voltage) were applied to the PS film.

To produce Si nanopillars, the NCP PS template was first coated with a 15‑nm Au film via sputtering. Metal‑assisted chemical etching (MACE) was then performed by immersing the sample in a 5:1 (v/v) HF/H₂O₂ solution for 1 min. The resulting Si pillar arrays were characterized by FESEM.

Surface morphologies were examined by FEI Quanta 200 SEM and FEI Nova NanoSEM 450 for cross‑sectional imaging.

Results and Discussion

Figure 1a shows the pristine HCP arrangement of 100 nm PS spheres on Si. Defects such as multilayer stacking are visible, a common issue for sub‑200 nm spheres. Our focus on 100 nm spheres allows direct comparison with ICPE data and highlights the challenge of achieving sub‑100 nm features.

SEM images of PS nanoparticles after etching for 0 (a), 5 (b), 7 (c), 9 (d), 10 (e), and 11 min (f).

Using a 3 mA beam current and 1 kV voltage, we monitored diameter evolution over time. At 5, 7, and 9 min the diameters averaged 88 ± 9 nm, 75 ± 8 nm, and 54 ± 8 nm, respectively. After 10 min, the size uniformity decreased and diameters dropped to 34 ± 10 nm; 11 min left only a few residual particles.

Figure 2 plots the measured diameters versus time, revealing a nonlinear decrease characteristic of anisotropic etching. The cross‑sectional images in Figure 3 confirm a transition from spherical to elongated shapes due to preferential top‑side sputtering. The longitudinal etch rate, derived from the model D = √{4R₀² – k²t²}, yields k ≈ 9.2 nm min⁻¹, consistent with the slow, controllable rates reported for ICPE.

Time dependence of diameter reduction. Experimental data (dots) align with the theoretical curve (red) using k = 9.2 nm min⁻¹.

Cross‑sectional FESEM of 200 nm PS spheres (a) and 5‑min etched spheres (b). Shape transition is evident.

At extended etching times, the rate accelerates, likely due to cumulative thermal energy from ion bombardment. The temperature rise, while remaining below the melting threshold of PS (~135 °C), enhances physical sputtering and contributes to the observed rate increase [27–29].

To assess particle adhesion, we compared IBE‑fabricated NCP arrays with ICPE‑produced ones. After 3 min HF etching, ICPE particles detached while IBE particles remained firmly attached, underscoring the superior fastness imparted by the ion beam process. Removal of PS residues is achievable with dichloromethane (2 h). These results suggest that IBE‑derived NCP arrays are commercially viable and can be readily integrated into nanofabrication workflows.

Beam current dependence was explored (3–10 mA). Higher currents increased the etch rate, reaching 18.9 nm min⁻¹ at 9 mA, but also induced surface roughening of the Si substrate (Fig. 4d). Beam voltage effects were modest; diameters decreased slightly from 500 V to 1 kV, with saturation beyond 1 kV (Fig. 6). This indicates that kinetic energy plays a secondary role compared to ion flux.

SEM of PS nanoparticles etched for 5 min at 5, 7, 9, and 10 mA.

Etch rate versus beam current.

Average diameter versus beam voltage (3 mA, 5 min).

Finally, the IBE‑derived NCP PS template was used to fabricate Si nanopillar arrays via MACE. FESEM imaging (Figure 7) shows uniform pillars (~54 nm diameter, ~100 nm height) with PS residues still present at the tops.

FESEM of Si nanopillar arrays.

Conclusions

By exposing a monolayer of 100 nm PS spheres to Ar⁺ ion beam, we achieved hexagonal NCP arrays with diameters tunable between 34 nm and 88 nm. Etching kinetics were governed by ion current (9.2–18.9 nm min⁻¹) and exhibited a time‑dependent acceleration due to thermal effects. The resulting NCP template enabled the fabrication of ordered Si nanopillars (~54 nm diameter, ~100 nm height) via metal‑assisted chemical etching. Compared with ICPE, the IBE process yields stronger particle adhesion and offers a scalable route for sub‑100 nm nanostructures, highlighting its potential for commercial nanosphere lithography.

Abbreviations

FESEM:

Field emission scanning electron microscope

IBE:

Ion beam etching

ICPE:

Inductively coupled plasma etching

PE:

Plasma etching

PS:

Polystyrene

RIE:

Reactive ion etching

SEM:

Scanning electron microscope