Automatic Shimming Method Using Compensation of Magnetic Susceptibilities and Adaptive Simplex for Low-Field NMR

Abstract

During the past several decades, inexpensive compact nuclear magnetic resonance (NMR) instruments have been widely used for on-site detections of chemical identity and sample analyses in industrial applications. In general, via shim coils and control strategies, automatic search shimming methods are capable of improving the magnetic field uniformity so that optimized signals can be acquired and particularly suited for rapid detections in compact NMR instruments. However, because these methods inherently endure the inefficiency in multidimensional shims and the sensitivity of starting points, it is time-consuming to obtain acceptable results in a field infected with distortions and strays. Here, two remedies that improve the shimming speed for the search shimming method are proposed. First, analyzing and compensating for magnetic susceptibility effects, we relocate starting points closely to optima and reduce the complexity of global exploration. Second, modifying the simplex movement with adaptive scale factors and gradient-like centroid, we have improved the convergence of the simplex shimming algorithm in a multidimensional search space. Via theoretical analyses and lab experiments, we find that the proposed shimming method outperforms regular shimming methods in terms of handling complex magnetic fields and reducing shimming times and hope that our study can help further improve NMR shimming methods in the future.