Laser-induced disassembly of a graphene single crystal into a nanocrystalline network

We report about investigations of time-dependent structural modifications in single-crystal graphene due to laser irradiation even at moderate power levels of 1 mW in a diffraction-limited spot. The modifications have been characterized by in situ scanning confocal Raman spectroscopy, atomic force height microscopy, and transport studies. The time evolution of the Raman spectrum reveals two different effects: on a short-time scale, dopants, initially present on the flake, are removed. The longer time scale behavior points to a laser induced gradual local decomposition of single-crystal graphene into a network of interconnected nanocrystallites with a characteristic length scale of approximately 10 nm due to bond breaking. The broken bonds offer additional docking sites for adsorbates as confirmed in transport and AFM height studies. These controlled structural modifications may for instance be valuable for enhancing the local reactivity, trimming graphene based gas sensors and generating spatially varying doping patterns.

Figure 1

(Color) Laser-induced change in the Raman spectrum. The black line (offset in intensity for clarity) is the spectrum of a pristine graphene monolayer. Other Raman spectra were obtained after laser exposure with 1 mW for the time interval indicated near each trace. Comparing the Raman spectrum after 18 h laser exposure with the spectrum of the pristine flake reveals dramatic changes. The appearance of the $D$ peak as well as the decrease in the $G$ and $2D$ peak intensities are signatures of structural modifications.

Figure 2

(Color) Time evolution of the $G$, $D$, and $2D$ Raman peaks. Panels (a), (e), and (g) show the intensity of these peaks as a function of time during laser exposure with 1 mW of power and a wavelength of 488 nm. For each data point on these curves a Raman spectrum was acquired in the range from 1024 to $3770{\text{cm}}^{-1}$ with a 30 s integration time. In panel (e) also the intensity ratio between the $D$ and $G$ peaks has been included as well as the intensity ratio between the $2D$ and $G$ peak [inset in panel (g)]. Panels (b), (f), and (h) display the time development of the position of each of these Raman peaks. The FWHM of the $G$ peak is given in (d). The observed changes with time suggest a structural modification of a single crystal of graphene into a network of graphene nanocrystallites as depicted in a cartoonlike fashion in panel (c).

Figure 3

(Color) Influence of removing dopants on the Raman spectrum of graphene. (a) field-effect curve, (b) $G$ Raman peak, and (c) $2D$ Raman peak of the as prepared flake (black solid lines), after evacuating the sample for 90 min (red dashed lines) and after heating the sample to $110°\text{C}$ in vacuum (blue dotted trace). Removal of dopants causes a redshift of the $G$ as well as $2D$ peak.

Figure 4

(Color) Time evolution of the $G$, $D$, and $2D$ Raman peaks under laser exposure. Dopants have been removed by heating and pumping. Compared to Fig. 2 all features attributed to the laser induced removal of dopants are absent. The time evolution points to the disassembly of single-crystal graphene into a network of graphene nanocrystallites.

Figure 5

(Color) Intensity maps of the $D$, $G$, and $2D$ Raman peaks before (left panels) and after laser exposure (right panels). (a) On the pristine flake there are virtually no defects and hence the $D$ peak is not observable. The white dashed line is a guide to the eyes and corresponds to the border of the graphene sheet where a few defects are located. (b) Inside a circle with a diameter of approximately $1.5\mu \text{m}$, the intensity of the $D$ peak is irreversibly enhanced after laser exposure. The size and position of the spot with a large $D$ count coincide with the laser spot. (c) Before the laser treatment, the intensity of the $G$ peak is nearly identical across the entire graphene sheet. It confirms the high crystalline quality of the flake. (d) After laser treatment, the intensity of the $G$ peak is reduced.

Figure 6

(Color) AFM image of a graphene monolayer before and after laser exposure recorded in tapping mode. (a) AFM height image of a pristine graphene flake. The top panel shows a height scan along the dashed line. The height $H$ of the substrate and the graphene surface are marked by the dashed black and green lines, respectively. The measured height difference is approximately 1 nm. (b) AFM height image after exposing one spot on the sample for 14 h with a 1 mW defocused 488 nm laser beam. An elevated area has appeared where the laser spot was positioned. In the top panel, the red dashed line marks its height. The difference with respect to the pristine graphene surface is approximately 1.5 nm. The diameter of the modified region coincides with the diameter of about $1.5\mu \text{m}$ of the defocused laser spot.

Figure 7

(Color) AFM measurement on a graphene flake on which multiple laser spots with different exposure time have been written. (a) Height image of the laser treated areas. Laser treated areas are numbered. A large number corresponds to a longer exposure time. (b) Analysis of the diameter of the laser treated spots in (a). The modification of graphene occurs first in the middle of the laser spot where the laser intensity has its maximum. (c) Analysis of the height as a function of the exposure time. After a small decrease in height (removal of dopants) the height grows until it reaches a maximum. The increased height is attributed to the attachment of ambient species after the laser has been turned off.

Figure 8

(Color) AFM image of a flake which has been treated by laser light at the edge. Only the graphene itself shows an elevated height. For the sake of comparison also a spot located entirely within the flake has been “written.”

Figure 9

(Color) Field-effect curves recorded on a graphene flake which has been exposed to laser irradiation. (a) AFM image of a typical device. (b) Field-effect characteristics for different laser exposure times. Before recording this transport data, the laser was switched off for 60 s. The removal of the dopants on the as prepared flake causes an initial shift of the charge neutrality point to lower back voltages, i.e., drop in the carrier concentration (first 20 min). As $s{p}^2$ bonds get cracked, the reactivity of graphene is locally enhanced and more dopants from the ambient can adsorb after the laser light has been switched off. This results in an increase of $p$-type doping.