Superconducting quantum interference devices (SQUIDs) are the most sensitive detectors of magnetic flux and have been used for decades in increasingly diverse fields of application, in order to measure tiny signals. A 'direct current' (dc) SQUID consists of a superconducting loop (with inductance L), which is interrupted by two Josephson junctions. The change in magnetic flux through the SQUID ring is detected as a change in voltage across the Josephson junctions. The sensitivity of a SQUID is characterized by the root mean square' (rms) noise power spectral density SΦ1/2 ("flux noise"). Typical thin film dc SQUIDs (with L ~ 10-100 pH) yield in magnetically shielded environment a flux noise of only a few µΦ0/Hz1/2 (Φ0 is the magnetic flux quantum).
Due to various fields of application in medical technology, magnetic data storage, ferrofluidics or quantum information processing, the interest in small spin systems, such as magnetic nanoparticles, nanotubes, nanowires or molecular magnets has strongly increased in recent years. This requires the development of suitable sensors for the investigation of the magnetic properties of these systems. Conventional measurement methods only allow the study of the magnetic properties of large ensembles. The related coupling effects complicate the interpretation of measurement results. In addition, anisotropic properties are averaged out, and due to strong dispersion in the magnetic properties with particle size, only ensemble averages are accessible. The development of an understanding of the magnetic properties of nanoparticles on the basis of ensemble measurements is therefore very limited. This motivates the development of appropriate measurement methods for the investigation of individual magnetic nanosystems. Compared to some already established spectroscopic methods, SQUIDs offer the possibility of directly detecting the reversal of magnetization M of single particles in external magnetic fields H (i.e. tracing out M(H) hysteresis loops) with unprecedented sensitivity. The ultimate goal is the resolution of the flipping of single electron spins. The relevant "figure of merit" in such an application is the spin sensitivity Sµ1/2= SΦ1/2/Φµ, which is typically given in units of µB/Hz1/2 (µB is the Bohr magneton); Here, the coupling factor Φµ is the magnetic flux Φ per magnetic moment μ, which is coupled by a magnetic particle into the SQUID. A high spin sensitivity of SQUIDs thus requires on one hand the lowest possible flow noise of SQUIDs and on the other hand the highest possible coupling factor. These requirements can be met by miniaturizing the SQUIDs and by placing the magnetic nanoparticles as close as possible to the SQUID. For these reasons, since many years the development of so-called, 'nanoSQUIDs' has been pushed ahead by many groups worldwide. So far, Josephson junctions were mostly based on nanopatterned constrictions ('constriction junctions'). Although these are relatively easy to implement, they have a number of disadvantages, which substantially complicate the operation and readout of nanoSQUIDs, and the optimization of the spin sensitivity.
Our group has two alternative approaches to the development of sensitive nanoSQUIDs. On the one hand, we use conventional superconductor-normalconductor-superconductor (SNS) Josephson junctions, based on Nb thin films with HfTi as the normal conducting barrier, for the above mentioned applications. This is in close cooperation with the Quantum Electronics department of PTB Braunschweig and the Cryophysics and Spectrometry department PTB Berlin. On the other hand, grain boundary junctions from the high-transition-temperature cuprate superconductor YBa2Cu3O7 (YBCO) are used, since this material allows operation in very strong magnetic fields in the tesla range. This is particularly relevant for studies on molecular magnets, since the required switching fields for magnetization reversal increase up to the Tesla reange with decreasing particle size. It should be mentioned that so far most SQUIDs are operated in magnetically shielded environment or at most in the earth’s magnetic field (~ 60 μT). The behavior, and in particular the sensitivity of SQUIDs operated in the mT or even Tesla range are largely unexplored and are a focus of our work on nanoSQUIDs.
Publications of our group
Articles in Scientific Journals: