A Droplet‐Based High‐Throughput SERS Platform on a Droplet‐Guiding‐Track‐Engraved Superhydrophobic Substrate

Background

SERS amplifies Raman signals through the strong electromagnetic fields of noble metal nanostructures, and coupling it with microfluidics allows in situ monitoring of analytes. When the SERS-active nanostructures sit inside a closed microfluidic channel, however, the authors note that they are damaged and contaminated by the fluid, which makes detection imprecise and requires cleaning between samples. Open-channel devices based on water-repellent surfaces can move small droplets with little friction and without leaving residue, which the authors propose as a way to avoid these problems.

Design and approach

A guiding track 2 mm wide with a 0.3 mm step was milled into a 1 mm thick copper (Cu) plate. Immersing the plate in 3 × 10⁻³ M silver nitrate grows Ag dendrites by galvanic displacement, because the standard electrochemical potential of Cu (0.34 V) is lower than that of Ag (0.80 V), and a 1 × 10⁻³ M 1H,1H,2H,2H-perfluorodecanethiol (PFDT) coating for 1 h then lowers the surface energy. The hierarchical dendrites provide dense SERS hot spots and, together with PFDT, superhydrophobicity. On the slightly tilted plate, a droplet rolls down the groove and halts at the stopper; the next droplet collides with it and, under the right conditions, bounces the first droplet onward and takes its place at the laser spot.

How it was evaluated

The authors followed dendrite growth by scanning electron microscopy and examined the composition of the fully grown structures by X-ray diffraction and energy-dispersive spectroscopy, and measured water contact and sliding angles against galvanic reaction time. Droplet collisions were mapped against Weber number for stopper heights of 0.2–0.6 mm in 0.1 mm steps, with 1 µL droplets on a substrate tilted at 1.4°. SERS performance was measured with 1 µL droplets of R6G, Nile blue A (NBA) and malachite green (MG) using a 514 nm Ar+ laser (5 mW) or a 633 nm He–Ne laser (2 mW); substrates made with reaction times of 2, 3, 5, 7, 10, 15, 20 and 25 min were compared at the 1363 cm−1 R6G peak using 10⁻³ M R6G, and a superhydrophobic substrate without Ag nanostructures served as a SERS-inactive control.

Key results

After a 10 s reaction the PFDT-coated surface had a water contact angle of 113.2° and droplets did not roll off; the contact angle rose to 150° at 1 min and saturated after 4 min, while the sliding angle fell below 5° at 5 min and saturated near 1° after 10 min. After 10 min, hierarchical structures of Ag nanoparticles, branches and leaves had formed. Droplets rolled along the groove without loss at tilt angles below 2°, and static droplets in the 2 mm track kept contact angles above 150° without shape deformation.

Collisions fell into three types as the Weber number of the incoming droplet increased: stopping after coalescence, bouncing in which the first droplet moves on and the second stops, and moving after coalescence. The bouncing regime needed for sequential measurement covered the broadest range of Weber numbers with a 0.3 mm stopper, and more than 100 consecutive droplet manipulations were achieved under these conditions. Droplet loss under laser irradiation was about 2% for irradiation times below 5 s.

The R6G signal increased with reaction time from 2 to 5 min and decreased from 5 to 25 min, which the authors attribute to the loss of hot-spot density as dendrites grew larger and smoother; 5 min was chosen as optimal. On this substrate the main R6G peaks were still resolved at 10⁻⁶ M, no R6G peaks appeared on the SERS-inactive control, and the enhancement factor was 6.3 × 10⁵, which the authors describe as comparable to closed-channel SERS microfluidic devices.

Successive measurements with and without 10⁻⁶ M R6G droplets, fifteen repeated spectra of 10⁻⁵ M NBA, and alternating MG and NBA droplets measured at the same laser spot showed the characteristic peaks of each analyte without spectral interference from earlier droplets. Droplets of 1 µL were released by a syringe pump at 0.5 µL s−1 and replaced at the laser spot without cleaning, whereas closed-channel microfluidic SERS sensors typically operate at about 0.2 µL s−1. The introduction states detection down to 10⁻⁶ M for R6G, NBA and MG, while the NBA and MG measurements reported in the results and methods used 10⁻⁵ M.

Open questions

The authors note that low analyte concentrations are difficult to detect under droplet conditions. They propose extending the method with multi-inlet-multi-outlet guiding tracks and with coalescing collisions that mix droplets containing different reactants, for example for nanoparticle synthesis or hydrogel crosslinking.

How it connects to other research

The related Nature Photonics study uses a superhydrophobic surface to concentrate dilute molecules at plasmonic nanostructures, whereas this sensor keeps droplets intact and uses a groove and stopper to deliver and replace them one after another at a fixed laser spot.