Can laser-excited surface waves pattern graphene oxide fast?
Scanning a line-shaped femtosecond laser over graphene oxide both reduces it and writes a highly regular nanoscale grating over centimetre areas, far faster than point-by-point laser writing.
Source
High-speed femtosecond laser plasmonic lithography and reduction of graphene oxide for anisotropic photoresponse
Study at a glance
- Design
- Other — Lab experiment: a line-focused 800 nm femtosecond laser was scanned over spin-coated graphene oxide films, the resulting structures were characterised (SEM, AFM, Raman, EDS), modelled with an analytic TE-plasmon dispersion and FDTD, and made into a photodetector.
- N
- No sample count reported; results come from fabricated films and one multi-electrode device.
- Population
- Graphene oxide films about 140 nm thick on SiO2/Si substrates
- Outcome
- Grating period and orientation, degree of photoreduction, light absorption, conductivity anisotropy and photoresponsivity
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The authors spin-coated graphene oxide films onto silicon and scanned them with a femtosecond laser focused into a long thin line. They measured the resulting surface ripples with electron and atomic force microscopy and mapped chemical reduction with Raman and elemental analysis. They explained the pattern with a model in which a depth-graded reduction lets the film support transverse-electric surface plasmons that interfere with the incoming light, checked with FDTD simulations. Finally they put twelve electrodes on the patterned film to test direction-dependent conduction and light response.
What they found
The laser produced gratings with a period of about 680 nm over a 10 by 12 mm area, oriented parallel to the laser polarisation, unlike most laser-induced ripples. Ridges and valleys were reduced to different degrees, with Raman D/G ratios falling from about 1.067 in untreated film to about 0.953 and 0.928. The patterned film absorbed about 20% more light, conducted about 21.7 versus 8.1 microsiemens along versus across the grating, and showed a steady photoresponsivity of roughly 0.7 mA per watt, though with slow rise and fall times of several seconds.
The limits
What it doesn't show
The surface-plasmon explanation rests on a semi-phenomenological model with assumed parameters for the depth-graded permittivity, not a direct measurement of the plasmon wave. The photodetector is slow (seconds) and weakly responsive, so the device result is a proof of concept rather than a competitive detector. The authors note the process becomes irregular for films thinner than 50 nm, and they report no statistics on repeatability across many samples.
Key terms
- Graphene oxide (GO) and reduced GO
- Graphene carrying oxygen-containing groups; removing them (reduction) restores conductivity and light absorption.
- LIPSS
- Laser-induced periodic surface structures: regular ripples that form on a surface after repeated laser pulses.
- TE surface plasmon
- A surface wave of electrons with transverse-electric polarisation, possible here because the film's permittivity changes with depth.
- Raman D/G ratio
- The intensity ratio of disorder (D) and graphitic (G) Raman peaks; a smaller ratio indicates more reduction.
- Conductivity anisotropy
- Direction dependence of electrical conduction, here higher along the grating lines than across them.
Flashcards
0 of 9 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What period did the laser-induced gratings have?
Common questions
Why is the grating period slightly smaller than the laser wavelength?
The period is set by interference between the incoming light and the surface plasmon wave, whose wavelength differs a little from the laser's; the authors found it nearly independent of pulse number and fluence.
Why is this faster than direct laser writing?
The line focus covers a 10 mm strip at once and the surface-wave feedback copies the pattern forward as it scans, so the authors estimate it takes about 1/14,000 of the time for a centimetre sample.
Why is the detector's response so slow?
The authors suggest heating-based (bolometric and photothermoelectric) effects dominate, and residual oxygen groups and rough nanostructures hinder conduction.
More on Plasmonics