Insulating behavior of magnetic spots in proton-bombarded graphite

Kelvin probe force microscopy measurements on micrometer small magnetic spots produced by proton bombardment on bulk graphite reveal a charge transfer from the center of the spot to an external ring with potential variation on the order of 50 mV. The total charge in the spot is neutral. The results can be well understood in terms of practically unscreened potentials, an insulating property, although the nonbombarded, surrounding graphite region exhibits good conductance. Scanning transmission x-ray microscopy measurements on magnetic spots prepared on graphitic films reveal similar charge distribution. The insulating behavior is fundamental to understand the magnetism in graphite.

Figure 1

(Color online) (a) AFM pictures showing three spots produced at a temperature of 110 K with a fluence of $0.124\text{nC}/\mu {\text{m}}^2$ and a proton current of 1 nA on HOPG. (b) KPFM pictures at the three spots. (c) AFM line scans across a horizontal line through the spots. (d) Potential variations obtained from the KPFM line scans for the three spots of (b). The color of the symbol corresponds to the same color line of the corresponding spot in (b).

Figure 2

$(●)$: Number of defects per ion and per $\mu \text{m}$ depth interval according to SRIM2003 Monte Carlo simulations (Ref. 13) for 2.25 MeV protons on HOPG. The data are obtained assuming a displacement energy of 35 eV for Frenkel pairs in HOPG. Following SQUID results obtained from HOPG samples irradiated with several thousands of spots at similar fluences as used for the spots discussed in this work (Ref. 7) the ferromagnetic region lies between the surface to $\sim 5\mu \text{m}$ depth. The rest of the irradiated sample depth shows a paramagnetic contribution. The overlapping region between ferromagnetism and paramagnetism depends on the fluence used. It is expected that it shifts to the right decreasing the fluence.

Figure 3

(a) Magnetic moment to the third power vs applied field for two irradiated HOPG samples. The left $y$ scale applies to the data from a sample irradiated with $25600$ spots at 5 K and 300 K (Ref. 7) and the right $y$ axis to the data of a broad proton beam irradiated HOPG sample (Ref. 6). Note that the data points are taken from the difference between the data of the samples after-minus before-irradiation and for decreasing field only. (b) Magnetic moment to the second power vs applied field divided by the magnetic moment from the same data as in (a). The continuous lines are only a guide to the eyes.

Figure 4

(Color online) Potential variation through the spot #1 from its center. The dashed (green) line corresponds to the 3D unscreened potential according to Eq. (3) with parameter $\rho /2ϵ=64\text{mV}/\mu {\text{m}}^2$; the continuous line corresponds to the 3D screened potential according to Eq. (4) with parameter $\rho /2ϵ=30\text{mV}/\mu {\text{m}}^2$ and $\lambda =6.7\mu \text{m}$; the short dashed line (only for $r<1.3\mu \text{m}$) corresponds to the unscreened 2D potential according to Eq. (5) with parameter $\left(\rho h\right)/2ϵ=42\text{mV}/\mu \text{m}$ and the dotted (red) line to the 2D screened potential according to Eq. (6) with parameters $\left(\rho h\right)/2ϵ=48\text{mV}/\mu \text{m}$ and $\lambda =5.3\mu \text{m}$. The bottom inset shows the line scan data for the spot #2 and the continuous curve the fit following the 3D screened potential according to Eq. (4) with parameter $\rho /2ϵ=50\text{mV}/\mu {\text{m}}^2$ and $\lambda =3.6\mu \text{m}$. The top left inset shows a sketch of the assumed cylindrical charge distribution from the top. The radii that are taken from the experimental data and included in the equations are $r_0=1.3\mu \text{m}$ and $r_1=1.8\mu \text{m}$.

Figure 5

(a) Transmitted x-ray intensity absorption at the $\pi$-resonance through a magnetic spot produced on a thin carbon film (sample A in Ref. 8). (b) The same but at the $\sigma$-resonance. The images are normalized using an image acquired at a photon energy below the absorption resonance to cancel out any contributions due to the topography. To eliminate any magnetic contrast from magnetic circular dichroism the images shown here are the result of averaging over two images acquired with opposite polarization of the x rays.