Oxidation-Based Continuous Laser Writing in Vertical Nano-Crystalline Graphite Thin Films

Nano and femtosecond laser writing are becoming very popular techniques for patterning carbon-based materials, as they are single-step processes enabling the drawing of complex shapes without photoresist. However, pulsed laser writing requires costly laser sources and is known to cause damages to the surrounding material. By comparison, continuous-wave lasers are cheap, stable and provide energy at a more moderate rate. Here, we show that a continuous-wave laser may be used to pattern vertical nano-crystalline graphite thin films with very few macroscale defects. Moreover, a spatially resolved study of the impact of the annealing to the crystalline structure and to the oxygen ingress in the film is provided: amorphization, matter removal and high oxygen content at the center of the beam; sp2 clustering and low oxygen content at its periphery. These data strongly suggest that amorphization and matter removal are controlled by carbon oxidation. The simultaneous occurrence of oxidation and amorphization results in a unique evolution of the Raman spectra as a function of annealing time, with a decrease of the I(D)/I(G) values but an upshift of the G peak frequency.


Ref Year
Base material

S2 -Calculation of the penetration depth of 488 nm photons into amorphous carbon
Following the standard framework, to estimate the penetration depth of photons into amorphous carbon, we use the Beer-Lambert law 43 : where I is the light intensity at depth x, I 0 is the incident light intensity and α is the attenuation constant. Then, the penetration depth is defined as: It is the depth at which the remaining electromagnetic power P equals 13% of its initial value P 0 (87% of the power has been absorbed).
The attenuation constant α can be calculated from the expression:  So for temperatures ranging from 25 to 600°C, the penetration depth remains lower than 38 nm.

S3 -X-ray photo electron spectroscopy
X-ray photo electron spectroscopy spectra shown in Supplementary Figure Figure S1 (a)).
No significant difference between the annealed and the un-annealed material is detected (Supplementary Figure S1

S4 -Evolution of the G peak width as a function of time
The width of the G peak (Supplementary Figure S2) features a strong drop at 1,080s, suggesting a sudden increase in the fraction of more graphitic matter. This occurs at the same time as a sudden increase in the amount of reflected light (Supplementary Figure S3), which we assign to the removal of carbon at the center of the crater (uncovering of the Ti layer).
Hence, this result suggests that the remaining carbon (on the sides of the crater) is more graphitic than the carbon at the center.
Supplementary Figure S2 -Evolution of the G peak width as a function of time.

S5 -Evolution of the Rayleigh scattering maximum as a function of time
The TEM images show that in some cases, matter is removed until the Ti layer is directly exposed to the laser beam. As Ti is highly reflective, we expect to observe a jump in the reflected light when Ti first gets directly exposed to the light. The reflected light is monitored qualitatively by measuring the intensity of the Rayleigh scattering of the Raman signal, namely the intensity of the peak centered at 0 cm -1 in terms of Raman shift (elastic process).
In Supplementary Figure S3, we plot the values of the maximum of the Rayleigh peak as a function of time. It oscillates until 1,080 s, at which time the spectrometer becomes saturated.
This result supports the fact that the Ti layer is first fully exposed at time = 1,080 s. The oscillation in elastically reflected light before 1,080 s may be due to changes in the depth of the hole, which in turn change the amount of light that is trapped in the hole.
Supplementary Figure S3 -Evolution of the maximum values of the Rayleigh peak as a function of time.

S6 -Beam profile
We use a CMOS camera to obtain the power distribution of the beam of the 488 nm laser (Supplementary Figure S4). It is a simple Gaussian beam, with a maximum power at the center.
Supplementary Figure S4 -Beam profile of the 488 nm laser obtained with a THORLABS DCC1545M CMOS camera.

S7 -Fitting the Raman spectra and extracting data on the standard errors with Scilab
We fit Raman spectra with three Lorentzians for the T, D and G peaks and add a baseline. The corresponding Scilab 45 code is: Then, we use the Bootstrap method to calculate random subsets of data from the initial data set. These subsets are then also fitted by three Lorentzians and a baseline, and the extracted