Diamond magnetometry of superconducting thin films

In recent years, diamond magnetometers based on the nitrogen-vacancy (NV) center have been of considerable interest for applications at the nanoscale. An interesting application which is well suited for NV centers is the study of nanoscale magnetic phenomena in superconducting materials. We employ NV centers to interrogate magnetic properties of a thin-layer yttrium barium copper oxide (YBCO) superconductor. Using fluorescence-microscopy methods and samples integrated with an NV sensor on a microchip, we measure the temperature of phase transition in the layer to be $70.0\left(2\right)$ K and the penetration field of vortices to be $46\left(4\right)$ G. We observe pinning of the vortices in the layer at 65 K and estimate their density after cooling the sample in a $\sim 10$-G field to be $0.45\left(1\right)$ $\mu$${\text{m}}^{-2}$. These measurements are done with a 10-nm-thick NV layer, so that high spatial resolution may be enabled in the future. Based on these results, we anticipate that this magnetometer could be useful for imaging the structure and dynamics of vortices. As an outlook, we present a fabrication method for a superconductor chip designed for this purpose.

Figure 1

(a) Level diagram of the NV center. ${}^3{A}_2$ and ${}^{3}E$ are the ground and excited triplet states, respectively. ${}^1{A}_1$ and ${}^{1}E$ are the intermediate singlet states involved in the optical-pumping process. The various spin levels are denoted by black lines. The radiative (nonradiative) transitions are denoted by solid (dashed) arrows. (b) The ODMR spectrum taken at different external magnetic fields applied perpendicular to the surface of the diamond. The field was generated with a single coil driven by the currents indicated in the legend. The diamond used in the experiment is cut along a (110) plane, and the green light beam is perpendicular to its surface. (c) The experimental setup. The sample area is shown enlarged at the bottom. The distance between the NV sensor and the outer side of the cryostat window is $\sim 2.3$ mm. Such a distance enables us to focus the green light onto the sensor, as the maximal working distance of the objective is 4 mm (SC, YBCO superconductor; MW, microwave)

Figure 2

Signal enhancement of the diamond sensor. (a) A comparison between the ODMR spectrum, before (blue line) and after (red line) the sample was annealed in oxygen. The contrast is enhanced due to conversion of ${\text{NV}}^0$ centers to ${\text{NV}}^{-}$. (b) An error signal (blue line) derived from the fluorescence signal (green line). The signal-to-noise ratio (SNR) is highly enhanced.

Figure 3

(a)–(e) Several ODMR signals taken at different temperatures with an applied external magnetic field of 15 G. The dashed black lines denote the Zeeman splitting, for each case. The up/down arrows indicate if the temperature was raised or lowered. (a) An ODMR signal taken below $T_c$, at 62 K. The system is in the Meissner state, and the external field does not penetrate the superconductor layer. (b) The signal at the phase transition. Here, we see a partial penetration of the field (a Zeeman splitting of 42 MHz corresponds to a $\sim 10$-G field in the $\widehat{z}$ direction). (c) This signal, at 74 K, indicates that the system is no longer in the superconducting phase. In this graph, we also demonstrate the fit to a derivative of a Lorentzian function, used to determine the zero crossing of the error signal. We use this fit on all ODMR signals to find the value of the Zeeman splitting. (d) Signal, after taking the temperature down again, without turning the external field off. Since the critical field at $T_c$ is zero, vortices penetrate the layer, leading to the average magnetic field which we measure. (e) As detailed in the text, at 60 K the magnetic field was turned off. Defects in the superconductor layer lead to flux pinning, evidenced by the field measured by the NV centers. (f) The phase-transition curve of the superconductor layer. Plotted is the Zeeman splitting between the $0\to 1$ and the $0\to -1$ resonances of the ${35}^{\circ }$ NV-axis alignments. The blue data points belong to the ascending temperature sequence, while the red ones to the descending sequence. The external field during the measurements was $\sim 15$ G excluding the last measurement (at $T=60$ K), where the field was turned off. We fit the blue data points to a sigmoid function (blue line), and extract a critical temperature of $T_c=70.0\pm 0.2$ K. The red line is to guide the eye.

Figure 4

(a), (b) ODMR spectra taken during the experimental sequence for finding the local penetration field $H_P$. The two different signals, taken with the same external field, reflect different responses of the material: (a) taken after increasing the field from zero, meaning the system at this spatial point is in the Meissner state, and the external field is screened; (b) recorded after lowering the field from $\sim 50$ G, which is above $H_P$, thus measuring the field of the pinned vortices; (c) the Zeeman splitting detected with the NV-diamond sensor during the measurement sequence. The applied magnetic field is proportional to the coil current. The plot demonstrates the local transition from the Meissner state to an intermediate state wherein vortices are in the sample. The blue points correspond to increasing coil current while the red points correspond to the decreasing-current sequence. The blue and the red lines are to guide the eye.

Figure 5

Different stages in the fabrication of the superconductor chip. (a) The smallest square on the chip after the first photoresist deposition. To realize a pattern on the MgO substrate, we used a positive mask and etched the YBCO with phosphoric acid. The cut corners are an artifact of the exposure process. (b) Following the YBCO etching, we covered the chip with a photoresist, after writing a negative mask. In this image, of one of the squares, the surface is covered with photoresist except for the desired pattern. A minor misalignment, of $\sim 2\mu$m, can be observed. (c) Image of the final chip, after covering it with silver, and lifting off the photoresist. The small square is marked with an arrow as it is not visible on this scale. (d) Zoomed image of two of the squares. The actual sizes, measured by microscope, proved to be somewhat smaller than the planned ones (25 and $75\mu$m).