Abstract:
The development of nano-sized direct current (dc) superconducting quantum interference devices (SQUIDs) is an intensive field of research, driven by the need for sensitive measurement techniques for the investigation of small spin systems ranging from the mTesla up to the Tesla range. This work gives a contribution to this field.
Both low-$T_c$ ($T_c$ is the transition temperature) SQUIDs and gradiometers based on Nb/HfTi/Nb superconductor/ normal metal/ superconductor (SNS) type Josephson junctions and high-$T_c$ dc SQUIDs based on $\rm{YBa}_2\rm{Cu}_3\rm{O}_7$ (YBCO) grain boundary junctions (GBJs) have been investigated. Since the rms spin sensitivity $S_\mu^{1/2}$ can be optimized by nano-patterning of the SQUIDs, advanced patterning techniques, i.e., electron-beam lithography and focused ion beam milling have been used in order to reduce the size of the SQUIDs, yielding linewidths well below $1\,\mu\rm{m}$. For both types of SQUIDs we implemented a galvanically coupled coil, where part of the SQUID loop is used to apply a magnetic flux without the need of an external coil.
The SQUIDs have been characterized with respect to their transport and noise characteristics at temperature $T=4.2\,\rm{K}$. In low magnetic fields ($\le 1\,\rm{mT}$), for the SNS type SQUIDs, a flux noise $S_\Phi^{1/2}=250\,\rm{n}\Phi_0/\rm{Hz^{1/2}}$ in the white noise regime and a maximum coupling factor $\phi_\mu\sim8.5\,\rm{n}\Phi_0/\mu_B$, calculated by a new method developed in this work, yielding a spin sensitivity $S_\mu^{1/2}=29\,\mu_B/\rm{Hz}^{1/2}$, was achieved. We demonstrated the reliable operation of such SQUIDs in magnetic fields up to $\sim 50\,\rm{mT}$. For higher fields we observed jumps in the voltage across the SQUID, which prevented a reliable operation. By combining the nanoSQUID with a low-temperature magnetic force microscope (LTMFM), we were able to correlate the jumps with the entrance of Abrikosov vortices into the superconducting leads of the SQUID. With this combined setup of a LTMFM and the nanoSQUID, for the first time we were able to measure the magnetic coupling $\Phi_\mu(\vec{r})$ between the magnetic moment $\vec{\mu}$ of a nano-magnet, i.e. a Ni nano-tube, and the SQUID loop as a function of the position $\vec{r}$ of the nano-tube. The results of this experiment were in good agreement with the numerical calculations of $\Phi(\vec{r})$ based on the London equations.
For the YBCO nanoSQUIDs we developed a fabrication technique based focused ion beam milling, capable to pattern GBJs with a width below $100\,\rm{nm}$ without significant degradation of the critical current density $j_c$.
The first generation of such nanoSQUIDs had a good $S_\Phi^{1/2}=4\,\mu\Phi_0/\rm{Hz}^{1/2}$ and maximum $\phi_\mu\approx10\,\rm{n}\Phi_0/\mu_B$, yielding a rms spin sensitivity $S_\mu^{1/2}=390\,\mu_B/\rm{Hz}^{1/2}$ at low magnetic fields. Subsequently improved SQUIDs with an galvanically coupled coil yielded $S_\mu^{1/2}=62\,\mu_B/\rm{Hz}^{1/2}$ in low magnetic fields and only slightly reduced $S_\mu^{1/2}=110\,\mu_B/\rm{Hz}^{1/2}$ even at high magnetic fields up to $\mu_0H=1\,\rm{T}$.
In parallel to the investigation of nanoSQUIDs, a second approach in order to improve the sensitivity of SQUIDs has been investigated. Numerical simulations revealed that a significant improvement of $S_\Phi^{1/2}$ up to a factor of 3 can be achieved by strong asymmetries in the resistance of the two Josephson junctions of a dc SQUID. The theoretical predictions have been investigated experimentally with SQUIDs based on a Nb/Al-AlO$_x$/Nb trilayer fabrication process using a combination of photolithography and Ar ion etching. An asymmetric SQUID where one junction was shunted with an external resistor $R_{\rm{shunt}}$ and the other junction was unshunted and a corresponding symmetric SQUID have been measured with respect to their transport and noise properties at $T=4.2\,\rm{K}$. The results of these measurements were in very good agreement with numerical simulation, confirming an improvement of $S_\Phi^{1/2}$ by a factor of $\sim3$ due to the resistance asymmetry. For nanoSQUIDs, this approach could further improve the spin sensitivity beyond technological and physical limitations of the miniaturization.