Indications for intrinsic superconductivity in highly oriented pyrolytic graphite

High-resolution magnetoresistance data in highly oriented pyrolytic graphite thin samples manifest nonhomogenous superconductivity with critical temperature $T_c\sim 25\text{K}$ and higher temperature. Our claim is based mainly in the observation of anomalous hysteresis loops of resistance versus magnetic field that cannot be assigned to magnetic irreversibility but indicates the existence of Josephson-coupled superconducting grains. In addition we observe quantum resonances that can be assigned to Andreev reflections and the absence of Schubnikov de Haas oscillations. The results indicate that graphite is a system with nonpercolative superconducting domains immersed in a semiconductinglike matrix. As possible origin of the superconductivity in graphite we discuss interior-gap superconductivity when two very different type of carriers with different masses are present.

Figure 1

Normalized resistance as a function of temperature at zero applied field for three HOPG samples. Samples (1) and (3) were obtained from the same HOPG grade ZYA (rocking curve width of $0.4°$ rocking curve width) bulk sample from Advanced Ceramics. Particle-induced x-ray emission indicates impurity levels of metallic elements below $5\mu \text{g}/\text{g}$ with exception of $V$ $\left(16\mu \text{g}/\text{g}\right)$. Sample characteristics: (1) $R\left(275\text{K}\right)=72\text{m}\Omega$, thickness $d\simeq 10\mu \text{m}$, length $l\simeq 4.4\text{mm}$, and width $w\simeq 1\text{mm}$; (2) $R\left(275\text{K}\right)=6.5\text{m}\Omega$, $d\simeq 30\mu \text{m}$, $l\sim 2\text{mm}$, and $w\simeq 0.9\text{mm}$; (3) $R\left(275\text{K}\right)=16\Omega$, $d\simeq 12\text{nm}$, $l\sim 30\mu \text{m}$, and $w\sim 15\mu \text{m}$ with $\sim 9\mu \text{m}$ distance between nearby voltage electrodes; see inset in Fig. 4a. The Pd electrodes (to avoid Schottky barriers) were prepared using conventional electron lithography. Sample (2) was obtained from HOPG grade ZYC bulk sample (rocking curve width $3.5°)$.

Figure 2

(Color online) Right $y$ axis: Resistance vs temperature for sample 4. The dimensions of this sample were $\sim 25\times 15\times 0.013\mu {\text{m}}^3$. The voltage electrode distance was $14\mu \text{m}$ for the measurement shown in this figure. Right $y$ axis: Amplitude $A$ in microvolt of the largest oscillation observed in the magnetoresistance with the largest field period. Inset: First derivative of the resistance vs $T$ below 100 K.

Figure 3

(Color online) Resistance vs temperature of sample 5 obtained at two different regions of the same sample. The thickness of this sample was 75 nm. Inset: optical microscope picture of sample 5 with the distribution of the Pd electrodes at different sample positions. The used voltage electrodes for the corresponding measured curve of the main panel are indicated. The input current electrodes are shown by arrows and the letter “I” in the sketch. The lateral dimensions of sample 5 can be obtained from the black bar at the bottom right of the inset.

Figure 4

(Color online) (a) Resistance of sample 3 between two adjacent voltage electrodes as a function of magnetic field. A close inspection of the MR of this sample at fields above 0.5 T reveals an anomalous behavior, namely, the MR oscillates. The oscillations shown in the insets were obtained after subtracting a quadratic field dependence around 0.3 and 6.4 T. These small-field-period oscillations in the resistance are superposed to oscillations of larger amplitude and field period; see inset in (b). Further measurements indicate that the overall shape, field positions, and period are independent of the field sweeping rate, field step, and field sweep direction. The oscillations are observed at low as well as high fields as expected because the slope of $R$ vs $B$ does not depend appreciably with field (Ref. 14). Different periods as well as oscillation amplitudes are observed for other samples. The inset shows an optical microscope picture of the sample with the Pd electrodes. Counting clockwise from input current electrode one at the right, the data shown were taken between electrodes 3 and 4 (Ch.2). (b) Temperature dependence of the voltage amplitude of the two oscillations taken from the Fourier fit; see inset. The continuous lines are guide. The inset shows the data at 2 K after subtraction of a linear field background and the continuous line is the Fourier fit with periods 0.1 and 0.387 T. These periods are independent of temperature within experimental resolution.

Figure 5

(Color online) (a) Normalized resistance vs temperature at different constant magnetic fields applied normal to the graphene layers for sample 6. This sample is similar to sample 4, see Fig. 2, with $\simeq 13\text{nm}$ thickness. The voltage was measured between the two electrodes separated by $14\mu \text{m}$ with the input current electrodes at the top and bottom; see inset. The sample at this voltage electrode position shows semiconducting behavior with resistances $R\left(100\text{K}\right)=477.8\Omega$ and $R\left(270\text{K}\right)=420.5\Omega$. Inset: optical microscope picture of sample 6 with the distribution of the Pd electrodes at different sample positions. (b) Similar to (a) but for sample 7 of 45 nm thickness and resistance $R\left(100\text{K}\right)=59.1\Omega$ and $R\left(265\text{K}\right)=57.7\Omega$. Inset: optical microscope picture of sample 7. The input current electrodes are at the top and bottom and the middle two are the voltage electrodes.

Figure 6

(Color online) (a) Strongly enhanced MR curve near zero field. A weak hysteresis appears similar to butterfly MR loops for superconductors with Josephson-coupled grains (Ref. 10). For a clear observation of the hysteresis we present in the inset the difference of the resistance curves (see text). (b) The height of the irreversibility maximum $\Delta R$ as well as their field positions $B_m$ (see inset) vs temperature. The continuous lines follow the function $\propto {\left[1-\left(T/11\right)\right]}^{1.5}$.

Figure 7

Sketch of an hysteretic loop for the resistance for two cases. The continuous arrows show the usual hysteretic loop of the resistance vs applied magnetic field for a material with ferromagnetic properties and a positive magnetoresistance. A similar loop is observed for homogeneous superconductors with conventional pinning of Abrikosov vortices. For the case of negative magnetoresistance the extrema are maxima instead of minima. The dashed arrows indicate the loop observed for granular superconductor with pinned Abrikosov vortices and Josephson fluxons at the intermediate regions. The position in field of the extrema are pinning dependent and in general for granular superconductors located nearer to zero field.