Signal to Noise Ratio of Single Domain Nanomagnet Amplification of Nuclear Magnetic Resonance Signals at Ambient Temperature

Abstract

Traditional inductive methods used for Magnetic Resonance Imaging (MRI) suffer from significant challenges due to the low signal-to-noise ratio (SNR) arising from the weak nuclear magnetic moment of protons when they precess about a large magnetic field at room temperature [1]. In an attempt to amplify the magnetic flux, it has been proposed to use magnetic nanoparticles of various geometrical shapes that are inserted into an NMR inductive coil [1-3]. Thus, the magnetic flux due to the response of the ensemble of proton spins to the RF field can be significantly enhanced as these nanoparticles have high transverse magnetic susceptibility under appropriate DC bias conditions. Prior work did not consider random thermal magnetization fluctuations in such nanomagnets that affect the SNR. In this work, we simulated the magnetization dynamics of a single-domain ellipsoidal nanomagnet driven by an AC field (HAC) along the easy axis for different bias fields (HDC) along the hard axis of the nanomagnet (Fig 1a). Importantly, our simulation also included the effect of room temperature thermal noise as shown in Fig 2. For HDC =hc, the energy profile is nearly flat (Fig 1b) that leads to the maximum magnetization amplification (Fig 1c) in response to the driving AC field that mimics the field due to proton spins precessing at nuclear magnetic resonance (NMR) frequency. However, the flat energy profile also increases the random magnetization fluctuations due to the thermal noise (Fig 2). Nevertheless, the SNRs for h=0, 0.5hc, 0.75hc, 1.0hc and 1.5hc are 0.003, 0.29, 0.63, 0.7 and 0.23 respectively showing h= hc provides the highest SNR. However, these are considering a single nanomagnet only. We will further investigate improvement of NMR detection SNR as a function of the coil-filling factor (millions of magnetic nanoparticles), nanoparticle shape, applied NMR polarizing field, and narrow band filtering to increase the SNR > 1000.