Overview

This study, co-authored by Jungmok Seo (then at the Center for Biomaterials, Korea Institute of Science and Technology), who now leads the BLISS Lab at Yonsei University, made a hydrogen (H2) sensor from arrays of slanted polyurethane acrylate (PUA) nanofibers coated with palladium (Pd) on one side, modeled on the slanted sensilla of insect antennae. In 4 vol% H2 in nitrogen, the Janus nanofibers bent toward the upright position and transmittance at 530 nm changed by about 65% with a response time of 5.1 s; with a polarizer the change reached about 80%, enough to read printed letters through the array by eye.

Background

Hydrogen is hazardous above about 4% by volume in air, and some conventional hydrogen sensors rely on electrical read-out through semiconductors, protonic conductors or platinum wires, which itself requires precautions against ignition. Optical sensors based on surface plasmon resonance need expensive analyzers, and earlier eye-readable chemochromic sensors such as WO3-Pd responded in about 10 min at room temperature and atmospheric pressure. The authors set out to make a flexible, simply fabricated sensor whose geometry changes visibly in hydrogen.

Design and approach

PUA nanofibers 400 nm in diameter and 4 µm tall, with a modulus of about 20 MPa, were moulded in a hexagonal array with 1.2 µm spacing, and Pd was sputtered onto one side at 45°, which also left the fibers slanted. When Pd absorbs hydrogen it forms palladium hydride and expands, so the Pd-coated side lengthens and the fiber straightens; when hydrogen is removed the fiber returns to its slant. Because slanted fibers shade the substrate more than upright ones, this motion acts as an optical shutter, and the uniform bending direction also makes the array polarizing. Arrays of 2 × 2 cm2 were made on flexible film.

How it was evaluated

The authors compared fibers 100 or 400 nm in diameter coated with Pd of different thickness against flat Pd films on PUA, and fibers 3 µm in diameter and 18 µm tall. Bending was imaged in cross section under 10 vol% H2 and modeled by beam-bending theory and finite-element simulation. Transmittance at 530 nm, where the eye is most sensitive, was measured in a gas-tight chamber at room temperature and atmospheric pressure for 0.1–10 vol% H2 in nitrogen, with response time defined as the time for a 10% to 90% change, along with cycling, polarization extinction ratio (PER) and read-out through a commercial polarizer. Wetting and switchable adhesion were tested with 4 vol% H2.

Key results

Fibers 400 nm in diameter with 100 nm of Pd gave the largest transmittance change, about 65% in 4 vol% H2, while 100 nm fibers, spaced 300 nm apart and thus below the 530 nm measuring wavelength, performed similarly to flat Pd, and 3 µm fibers did not bend after Pd deposition or in hydrogen. The fabricated thickness series is listed as 0, 40, 80 and 120 nm, whereas the best-performing array and Figure 1d are described with a 100 nm Pd coat. In 10 vol% H2, the bending angle decreased from about 17.3° to about 10.2° within 4 s and recovered within 6 s after hydrogen removal, consistent with the model estimate of about 7.7° for Pd strain below 3.5%.

The array responded reversibly from 0.1 to 10 vol% H2, with a sharp increase in bending and transmittance near 1 vol% that the authors attribute to the α-to-β phase transition of palladium hydride. Its optical sensitivity was significantly higher than that of the flat Pd film, the response remained stable over repeated cycles (reported as <40 cycles) without notable degradation, and the response time fell from 17.6 to 3.6 s over 0–7 vol% H2; at 4 vol% it was 5.1 s, faster than the flat Pd layer.

The aligned fibers gave a PER of about 4.5, compared with nearly zero for flat Pd, and a polarizer placed perpendicular to the bending direction raised the sensitivity at 4 vol% H2 from 65% to about 80%. In the eye-readable demonstration, printed letters beneath the array became clearly legible when the chamber was filled with 4 vol% H2, whereas without the polarizer the change was barely visible.

Hydrogen exposure changed the critical wetting angle toward one side from 58° to 64°, above the intrinsic contact angle of about 62°, so that liquid spreading switched from one direction to both. For interlocked arrays of 100 nm fibers with 40 nm of Pd, shear adhesion in air was 5.9 and 17.9 N cm−2 in the two directions, with a hysteresis of about 3.0, and changed to 7.9 and 14.3 N cm−2, with a hysteresis of about 1.8, after hydrogen exposure. Two interlocked 1 × 1 cm2 arrays held a 1.6 kg object (15.7 N) in air and released it when 4 vol% H2 was blown onto them.

Related research

RESEARCH NEWS

Research in the news

Unverified candidates and access limits (1)

Sources used in this summary

This summary is based on the publisher PDF of the article and its Supporting Information PDF, including the methods and figure legends; the supplementary movies were not reviewed.

DOI: 10.1002/adfm.201701618