Tracking energy scale variations from scan to scan in nuclear resonant vibrational spectroscopy: In situ correction using zero-energy position drifts ΔEi rather than making in situ calibration measurements.

Abstract

Nuclear resonant vibrational spectroscopy (NRVS) is an excellent modern vibrational spectroscopy, in particular, for revealing site-specific information inside complicated molecules, such as enzymes. There are two different concepts about the energy calibration for a beamline or a monochromator (including a high resolution monochromator): the absolute energy calibration and the practical energy calibration. While the former pursues an as-fine-as-possible and as-repeatable-as-possible result, the latter includes the environment influenced variation from scan to scan, which often needs an in situ calibration measurement to track. However, an in situ measurement often shares a weak beam intensity and therefore has a noisy NRVS spectrum at the calibration sample location, not leading to a better energy calibration/correction in most cases. NRVS users for a long time have noticed that there are energy drifts in the vibrational spectra's zero-energy positions from scan to scan (ΔEi), but their trend has not been explored and utilized in the past. In this publication, after providing a brief introduction to the critical issue(s) in practical NRVS energy calibrations, we have evaluated the trend and the mechanism for these zero-energy drifts (ΔEi) and explored their link to the energy scales (αi) from scan to scan. Via detailed analyses, we have established a new stepwise procedure for carrying out practical energy calibrations, which includes the correction for the scan-dependent energy variations using ΔEi values rather than running additional in situ calibration measurements. We also proved that one additional instrument-fixed scaling constant (α0) exists to convert such "calibrated" energy axis (E') to the real energy axis (Ereal). The "calibrated" real energy axis (Ereal) has a preliminary error bar of ±0.1% (the 2σE divided by the vibrational energy position), which is 4-8 times better than that from the current practical energy calibration procedure.