Nanoscale Printing of Indium-Tin-Oxide by Femtosecond Laser Pulses

Abstract

For constructing optical and electrical micro-devices, the deposition/printing of materials with sub-1 μm precision and size (cross-section) is required. Crystalline c-ITO (indium tin oxide) nanostructures were patterned on glass with sufficient precision to form 20-50 nm gaps between individual disks or lines of ∼250 nm diameter or width. The absorbed energy density [J/cm3] followed a second-order dependence on pulse energy. This facilitated high-resolution and precise nanoscale laser-writing at a laser wavelength of 515 nm. Patterns for optical elements such as circular gratings and micro-disks were laser-printed using ITO as a resist. Unexposed amorphous a-ITO was chemically removed in aqueous 1% vol. HF solution. This use of a-ITO as a solid resist holds promise for metamaterial and micro-optical applications.

Keywords: IR; ITO; laser printing; solid resist.

Figure A1

FDTD calculation of light-intensity distribution (within $1\mathsf{\mu }$m${}^3$ volume) for an array of ITO nano-disk pairs; period in XY-plane is ${\mathrm{\Lambda }}_{x,y}=1\mathsf{\mu }$m, diameter of ITO disk $2r=250$ nm, the gap $\mathrm{\Delta }x=40$ nm, height of disk $h=80$ nm. Light source is a plane wave with a normalised E-field ${\left|E\right|}^2=1$. Glass is defined as dielectric with refractive index $n=1.4$ and the complex refractive index of ITO [36] was added to Lumerical material database as a table. Top-row is the side-view and bottom-row is the top-view E-field enhancement maps.

Figure 1

SEM images of ITO structures by nano-printing. (a) Two pulse-burst exposure of ITO film after wet etching in 1% vol. HF solution for 5 s. The ITO was printed using two laser-pulse bursts (515 nm/280 fs/10 pulses per burst/ pulse energy $E_p=0.92\pm 0.02$ nJ (on sample)) at separations $S_p$ varying from 0 to 300 nm. Focusing was performed using an $NA=0.9$ objective lens; polarisation linear (horizontal). Nanogaps can be controlled from 25 nm (b) to a size ten-times larger (c). Numerical aperture of the objective lens was $NA=0.9$ (focal spot diameter $\oslash =1.22\lambda /NA=698$ nm), thickness of ITO films $d\approx 60$ nm. Polarisation of laser pulses (b,c) was linear (vertical).

Figure 2

(a) Plot showing the length of two pulse-burst-exposed ITO nano-disks vs. separation between these two bursts S${}_p$ measured via SEM imaging. (b) AFM profilometry of laser-printed nano-disks. Two pulse-burst exposure of ITO film after wet etching in HF solution. Polarisation of laser E-field was horizontal, $NA=0.9$, pulse energy $E_p=0.92\pm 0.02$ nJ (on sample), $N=10$ pulse bursts per disk at $f=5$ kHz (for the two-spot exposures).

Figure 3

(a) Plot showing nano-disk diameter as a function of pulse number N measured by SEM for pulse energies between $E_p=0.77$ nJ and 0.46 nJ. (b) $E_p$ vs. N diagram for nano-disk fabrication: threshold of disk formation (×), region where different-sized disks can be made (○) and where size (diameter) saturates (▽). (c,d) c-ITO transitioning from single dots to a line on glass, recorded at different pulse energies $E_p$ for $N=10$ per irradiation site. (e) Formation of lines of different widths by selection of $E_p,N$, and separation $S_p$ along the line. (f) Pattern of separate lines using vertical Y shift. Polarisation linear (horizontal; along the line).

Figure 4

Plot showing the evolution of line-width vs. pulse energy $E_p$ for $N=10$ pulses per site at different separations $S_p$ along the line (a log–log plot); polarisation of E-field was linear (along the line) shown in the inset. The trend of dependence closely follows a nonlinear $\propto E_p^2$ dependence.

Figure 5

(a) SEM image of a Bessel-beam generating a micro-optical element recorded on glass. Polarisation of laser E-field was horizontal, $NA=0.9$, scanning speed $v_s=30\mathsf{\mu }$m/s and pulse energy $E_p=0.9$ nJ (on sample). (b) AFM view of the concentric grating with ITO lines that were 60 nm in height. (c) Bulk c-ITO recorded on glass. (d) AFM view of the bulk of ITO.

Figure 6

X-ray energy dispersion spectroscopy (EDS) analysis of c-ITO structure. (a) SEM image of c-ITO line structure after laser exposure and development. (b) Distribution of Si ($K_{\alpha }=1.739$ keV), O ($K_{\alpha }=0.525$ keV), In, Sn according their characteristic lines; 200 nm scale bar is same for all elemental maps. (c) Compositional maps of the four elements on the surface of sample. Inset shows SEM image of the mapped surface; all same 1 $\mathsf{\mu }$m scale bar. (d) The EDS spectra of pristine, laser irradiated and HF-developed ITO structures.

Figure 7

Numerical modeling of light intensity on an ITO metasurface consisting of nano-disk pairs. (a) Permittivity $\tilde{\epsilon }\equiv {\epsilon }_1+i{\epsilon }_2\equiv {(n+ik)}^2$ of ITO [36]. The epsilon-near-zero (ENZ) region $0<{\epsilon }_1<1$ is the near-IR spectral region. (b) Cross-sections of scattering, absorption and extinction ${\sigma }_{ex}\equiv {\sigma }_{ab}+{\sigma }_{sc}$ for a pair of ITO nanodisks calculated for 250 nm and 230 nm diameters using a total-field scattered-field (TFSF) source in Lumerical. Refractive index of glass $n=1.4$, the height of disk $h=80$ nm, gap 40 nm. Insets show characteristic light enhancement maps at different cross-sections and wavelengths. Linear polarisation of light was aligned perpendicular to the gap.