[Paper Review] Low-Noise YBa$_2$Cu$_3$O$_7$ NanoSQUIDs for Performing Magnetization-Reversal Measurements on Magnetic Nanoparticles

Tobias Schwarz, Roman Wölbing+7 more|arXiv (Cornell University)|Mar 20, 2015

Physics of Superconductivity and Magnetism5 citations

TL;DR

This paper presents low-noise YBa₂Cu₃O₇ (YBCO) dc nanoSQUIDs with grain boundary Josephson junctions, fabricated via focused ion beam patterning, achieving a white flux noise level below 50 nΦ₀/Hz¹/² at 4.2 K. The devices enable high-sensitivity magnetization-reversal measurements on magnetic nanoparticles, demonstrated by detecting 82.5 mΦ₀ flux changes from a 39 nm Fe nanowire switching via curling mode, with spin sensitivity of 3.7 μB/Hz¹/².

ABSTRACT

We fabricated YBa$_2$Cu$_3$O$_7$ (YBCO) direct current (dc) nano superconducting quantum interference devices (nanoSQUIDs) based on grain boundary Josephson junctions by focused ion beam patterning. Characterization of electric transport and noise properties at 4.2$\\,$K in magnetically shielded environment yields a very small inductance $L$ of a few pH for an optimized device geometry. This in turn results in very low values of flux noise $<50\\,{\ m n}\\Phi_0/{\ m Hz}^{1/2}$ in the thermal white noise limit, which yields spin sensitivities of a few $\\mu_{\ m B}/{\ m Hz}^{1/2}$ ($\\Phi_0$ is the magnetic flux quantum and $\\mu_{\ m B}$ is the Bohr magneton). We observe frequency-dependent excess noise up to 7$\\,$MHz, which can only partially be eliminated by bias reversal readout. This indicates the presence of fluctuators of unknown origin, possibly related to defect-induced spins in the SrTiO$_3$ substrate. We demonstrate the potential of using YBCO nanoSQUIDs for the investigation of small spin systems, by placing a 39$\\,$nm diameter Fe nanowire, encapsulated in a carbon nanotube, on top of a non-optimized YBCO nanoSQUID and by measuring the magnetization reversal of the Fe nanowire via the change of magnetic flux coupled to the nanoSQUID. The measured flux signals upon magnetization reversal of the Fe nanowire are in very good agreement with estimated values, and the determined switching fields indicate magnetization reversal of the nanowire via curling mode.

Motivation & Objective

• To develop ultra-low-noise nanoSQUIDs for high-sensitivity detection of magnetic nanoparticle (MNP) magnetization reversal.
• To overcome limitations of conventional SQUIDs in high magnetic fields and low-frequency noise, particularly 1/f noise.
• To demonstrate the feasibility of YBCO-based nanoSQUIDs for quantitative measurement of MNP magnetic switching dynamics.
• To identify and characterize sources of excess noise, such as defect-induced spin fluctuations in the SrTiO₃ substrate.
• To validate the device performance by measuring magnetization reversal in a 39 nm Fe nanowire encapsulated in a carbon nanotube.

Proposed method

• Fabricated YBCO dc nanoSQUIDs using focused ion beam patterning on a SrTiO₃ substrate with grain boundary Josephson junctions.
• Optimized device geometry to achieve a very low inductance L of a few pH, minimizing flux noise.
• Performed electrical transport and noise characterization at 4.2 K in a magnetically shielded environment.
• Employed numerical simulations based on London theory to determine the magnetic flux coupling factor φμ for a point-like magnetic moment at 10 nm distance.
• Applied a bias reversal readout scheme to suppress 1/f noise, though not fully eliminating frequency-dependent excess noise.
• Placed a 39 nm Fe nanowire on top of a non-optimized nanoSQUID and measured its magnetization reversal under in-plane magnetic fields.

Experimental results

Research questions

• RQ1 Can YBCO nanoSQUIDs with grain boundary junctions achieve sub-50 nΦ₀/Hz¹/² white flux noise at 4.2 K in a magnetically shielded environment?
• RQ2 To what extent does frequency-dependent excess noise limit the performance of low-inductance YBCO nanoSQUIDs, and can it be mitigated by bias reversal?
• RQ3 Does the measured flux signal from a magnetic nanowire switching via curling mode agree quantitatively with theoretical predictions based on its magnetic moment and geometry?
• RQ4 What is the spin sensitivity of the nanoSQUID for a magnetic moment located 10 nm above the SQUID loop, and how does it compare to existing technologies?
• RQ5 What is the origin of the observed MHz-band excess noise, and could it be linked to defect-induced spins in the SrTiO₃ substrate?

Key findings

• The optimized YBCO nanoSQUID achieved a white flux noise level of less than 50 nΦ₀/Hz¹/² at 4.2 K, corresponding to a spin sensitivity of 3.7 μB/Hz¹/² for a magnetic moment 10 nm above the SQUID loop.
• The measured flux change of ±82.5 mΦ₀ during magnetization reversal of the 39 nm Fe nanowire closely matched the theoretical estimate of 81.4 mΦ₀, confirming quantitative detection capability.
• The observed switching field of approximately ±100 mT is in excellent agreement with the predicted nucleation field of 103 mT for curling-mode reversal in a nanowire with dFe = 39 nm and λex = 5.8 nm.
• The frequency-dependent excess noise extending up to 7 MHz could only be partially suppressed by bias reversal, indicating the presence of additional noise sources such as defect-induced spin fluctuations in the SrTiO₃ substrate.
• The signal-to-noise ratio was improved by about one order of magnitude compared to micro-Hall magnetometry, with a noise amplitude of ~1 mΦ₀ versus ~10 mΦ₀ in prior measurements.
• The results confirm that YBCO nanoSQUIDs are suitable for detecting magnetization reversal in magnetic nanoparticles in moderate magnetic fields, with high sensitivity and quantitative accuracy.