The magnetic field noise in superconducting quantum interference devices (SQUIDs) used for biomagnetic research such as magnetoencephalography or ultra-low-field nuclear magnetic resonance is usually limited by instrumental dewar noise. We constructed a wideband, ultra-low noise system with a 45 mm diameter superconducting pick-up coil inductively coupled to a current sensor SQUID. Thermal noise in the liquid helium dewar is minimized by using aluminized polyester fabric as superinsulation and aluminum oxide strips as heat shields. With a magnetometer pick-up coil in the center of the Berlin magnetically shielded room 2 (BMSR2), a noise level of around 150 aT Hz^−1∕2 is achieved in the white noise regime between about 20 kHz and the system bandwidth of about 2.5 MHz. At lower frequencies, the resolution is limited by magnetic field noise arising from the walls of the shielded room. Modeling the BMSR2 as a closed cube with continuous μ-metal walls, we can quantitatively reproduce its measured field noise.

Very low residual magnetic field and field gradients are essential for a number of high resolution fundamental physical experiments and for further improvement of very sensitive magnetic measurement devices. The scope ranges from spin precession experiments, e.g. with 3He or neutrons, to biomagnetic measurements, like magnetoencephalograms, and to low field MR spectroscopy. One method of reducing environmental magnetic noise is to use a magnetically shielded room (MSR). Here, measures are demonstrated to improve residual field and field gradient inside a common MSR by a factor of more than 10 by a specific degaussing procedure, material selection of prefabricated parts and active shielding. The process is independent of the shielding factor and works also properly for heavily shielded rooms.

Magnetic nanoparticles (MNP) offer a large variety of promising applications in medicine thanks to their exciting physical properties, e.g., magnetic hyperthermia and magnetic drug targeting. For these applications, it is crucial to quantify the amount of MNP in their specific binding state. This information can be obtained by means of magnetorelaxometry (MRX), where the relaxation of previously aligned magnetic moments of MNP is measured. Current MRX with optically pumped magnetometers (OPM) is limited by OPM recovery time after the shut-off of the external magnetic field for MNP alignment, therewith preventing the detection of fast relaxing MNP. We present a setup for OPM-MRX measurements using a commercially available pulsed free-precession OPM, where the use of a high power pulsed pump laser in the sensor enables a system recovery time in the microsecond range. Besides, magnetometer raw data processing techniques for Larmor frequency analysis are proposed and compared in this paper. Due to the high bandwidth (≥100 kHz) and high dynamic range of our OPM, a software gradiometer in a compact enclosure allows for unshielded MRX measurements in a laboratory environment. When operated in the MRX mode with non-optimal pumping performance, the OPM shows an unshielded gradiometric noise floor of about $600 \mathrm{f}\mathrm{T}/\mathrm{c}\mathrm{m}/\sqrt{\mathrm{Hz}}$ for a $2.3 \mathrm{c}\mathrm{m}$ baseline. The noise floor is flat up to $1 \mathrm{kHz}$ and increases then linearly with the frequency. We demonstrate that quantitative unshielded MRX measurements of fast relaxing, water suspended MNP is possible with the novel OPM-MRX concept, confirmed by the accurately derived iron amount ratios of MNP samples. The detection limit of the current setup is about 1.37 μg of iron for a liquid BNF-MNP-sample (Bionized NanoFerrite) with a volume of 100 μL$$.

Non-invasive EEG detection of very high frequency somatosensory evoked potentials featuring frequencies up to and above 1 kHz has been recently reported. Here, we establish the detectability of such components by combined low-noise EEG/MEG. We recorded SEP/SEF simultaneously using median nerve stimulation in five healthy human subjects inside an electromagnetically shielded room, combining a low-noise EEG custom-made amplifier (4.7 nV/√Hz) and a custom-made single-channel low-noise MEG (0.5 fT/√Hz @ 1 kHz). Both, low-noise EEG and MEG revealed three spectrally distinct and temporally overlapping evoked components: N20 (<100 Hz), sigma-burst (450–750 Hz), and kappa-burst (850–1200 Hz). The two recording modalities showed similar relative scaling of signal amplitude in all three frequencies domains (EEG [10 nV] ≅ MEG [1 fT]). Pronounced waveform (peak-by-peak) overlap of EEG and MEG signals is observed in the sigma band, whereas in the kappa band overlap was only partial. A decreasing signal-to-noise ratio (SNR; calculated for n = 12.000 averages) from sigma to kappa components characterizes both, electric and magnetic field recordings: Sigma-band SNR was 12.9 ± 5.5/19.8 ± 12.6 for EEG/MEG, and kappa-band SNR at 3.77 ± 0.8/4.5 ± 2.9. High-frequency performance of a tailor-made MEG matches closely with simultaneously recorded low-noise EEG for the non-invasive detection of somatosensory evoked activity at and above 1 kHz. Thus, future multi-channel dual-mode low-noise technology could offer complementary views for source reconstruction of the neural generators underlying such high-frequency responses, and render neural high-frequency processes related to multi-unit spike discharges accessible in non-invasive recordings.

We present the prototype module of our extendible and robust multichannel SQUID magnetometer system. A large multi-module arrangement can be implemented by using up to 7 modules. The system is intended for high-precision measurements of biomagnetism and spin precession. Further demanding applications are magnetorelaxometry and ultra-low-field nuclear magnetic resonance (ULF NMR), where pulsed magnetic fields of up to 100 mT are typically applied. The system is operated inside the Berlin magnetically shielded room (BMSR-2) and equipped with 18 magnetometers consisting of niobium (Nb) wire-wound pick-up coils. A total of 16 small pick-up coils with 17.1 mm diameter form a regular grid with individual channels arranged to ensure system sensitivity covers all three orthogonal spatial directions. Two large hexagonal pick-up coils with an equivalent diameter of 74.5 mm sensitive in z-direction surround the grid at two different heights and are suitable for the detection of deep sources. Each pick-up coil is connected to the input of a thin-film Nb SQUID current sensor via a detachable superconducting contact. The SQUIDs are equipped with integrated input current limiters. Feedback into the pick-up coils is employed to minimise crosstalk between channels. The current sensor chip package includes a superconducting shield of Nb. The field distortion of the prototype and a multi-module arrangement was analysed by numerical simulation. The measured noise of the small magnetometers was between 0.6 and 1.5 fT , and well below 1 fT for the large ones. Using a software gradiometer, we achieved a minimum noise level of 0.54 fT . We performed ULF NMR experiments, verifying the system's robustness against pulsed fields, and magnetoencephalographgy (MEG) on somatosensory evoked neuronal activity. The low noise performance of our 18-channel prototype enabled the detection of high-frequency components at around 1 kHz by MEG.