Calibration Transfer of Partial Least Squares Regression Models between Desktop Nuclear Magnetic Resonance Spectrometers

Abstract

Low-field proton nuclear magnetic resonance (LF-1H NMR) devices based on permanent magnets are a promising analytical tool to be extensively applied to the process analytical chemistry scenario. To enhance its analytical applicability in samples where the spectral resolution is compromised, multivariate regression methods are required. However, building a robust calibration model, such as partial least squares (PLS) regression, is a laborious task because (1) the number of measurements required during the calibration process is large and (2) the procedure must be repeated when the instrument is changed or after a certain period due to the long-term stability of the instrument. Thus, the present work describes the application of calibration transfer methodologies (direct standardization (DS), piece-wise direct standardization (PDS), and double-window piece-wise direct standardization (DWPDS)) on LF-1H NMR to exempt the necessity of a recalibration procedure when moving from the original spectrometer to a second one with the same, lower, or higher magnetic field. These calibration transfer methodologies were tested with PLS models built on a 60 MHz (for the proton Larmor frequency) spectrometer to predict the specific gravity (SG), distillation temperature (T50%), and final boiling point (FBP) of commercial gasoline. The results showed that the DWPDS method applying only 2 to 7 transference samples enables the transference of all PLS models built on the primary instrument (60 MHz) to other (43, 60, and 80 MHz) different instruments, reaching the same RMSEP values as the primary instrument: 1.2 kg/m3 for SG, 5.1 °C for FBP, and 1.1 °C for T50%.

Parameter Visualization of Benchtop Nuclear Magnetic Resonance Spectra toward Food Process Monitoring

“…The possibility of NMR data transfer from a low field to a high field has also been investigated in time-domain NMR [44]. In addition, calibration transfer of low-field proton NMR ( 1 H-NMR) spectra using partial least squares regression in 40-, 60-, and 80-MHz benchtop NMR has been investigated [45].…”

A One-Dimensional U-Net-Based Calibration-Transfer Method for Low-Field Nuclear Magnetic Resonance Signals

“…Unfortunately, although LF-NMR spectral data are used to construct ML models with good predictive ability, the extent of the applicability of these models is dependent on instrument stability, even under the conditions of the same type of instrument. This is mainly attributed to the fact that the spectral changes related to the instrument itself and the external environment, such as magnetic field strength, magnetic susceptibility, and temperature, significantly affect the predictive ability of the model . It is well known that a change in the spectrum can alter key features of the same sample, which ultimately reduces the identification accuracy of the model.…”

“…In a two-dimensional (2D) diffusion-ordered spectroscopy study, Monakhova and Diehl used the PDS algorithm to reduce the RMSE of heparin from 647 to 513 Da . The double-window piece-wise direct standardization (DWPDS) algorithm was found to have the same effect on a PLS model of an NMR instrument with medium resolution . It is worth noting that all of the abovementioned algorithms involve solving an inverse matrix.…”

“…This is mainly attributed to the fact that the spectral changes related to the instrument itself and the external environment, such as magnetic field strength, magnetic susceptibility, and temperature, significantly affect the predictive ability of the model. 8 It is well known that a change in the spectrum can alter key features of the same sample, which ultimately reduces the identification accuracy of the model. Similarly, various spectroscopy devices also induce an instrument drift phenomenon (e.g., the electronic nose, 9 nearinfrared, 10 liquid chromatography-mass spectrometer, 11 and Raman spectrometer 12 ).…”

“…19 The double-window piecewise direct standardization (DWPDS) algorithm was found to have the same effect on a PLS model of an NMR instrument with medium resolution. 8 It is worth noting that all of the abovementioned algorithms involve solving an inverse matrix. However, when the matrix is not full rank, there will be a certain degree of error between the pseudo-inverse solution and the real solution.…”