Optimizing Data Processing for Nanodiamond Based Relaxometry

Abstract

The nitrogen-vacancy (NV) center in diamond is a powerful and versatile quantum sensor for diverse quantities. In particular, relaxometry (or T1), can be used to detect magnetic noise at the nanoscale. For experiments with single NV centers the analysis of the data is well established. However, due to relatively low brightness and reproducibility it is beneficial for biological experiments to use ensembles. While increasing the number of NV centers in a nanodiamond leads to more signal, a standardized method to extract information from relaxometry experiments is still missing. This article uses T1 relaxation curves acquired at different concentrations of gadolinium ions to calibrate and optimize the entire data processing flow, from the acquired raw data to the extracted T1. In particular, a bootstrap is used to derive a signal to noise ratio (SNR) that can be quantitatively compared from one method to another. At first, T1 curves are extracted from photoluminescence pulses. This work compares integrating their signal through an optimized window as performed conventionally, to fitting a known function on it. Fitting the decaying T1 curves leads to the relevant T1 value. This work compares here the three most commonly used fit models that are, single, bi, and stretched exponential. This work finally investigates the effect of the bootstrap itself on the precision of the result as well as the use of a rolling window to increase time resolution.