Extracting accurate infrared lineshapes from weak vibrational probes at low concentrations

Abstract

Fourier-transform infrared (FTIR) spectroscopy using vibrational probes is an ideal tool to detect changes in structure and local environments within biological molecules. However, challenges arise when dealing with weak infrared probes, such as thiocyanates, due to their inherent low signal strengths and overlap with solvent bands. In this protocol we demonstrate:•A streamlined approach for the precise extraction of weak infrared absorption lineshapes from a strong solvent background.•A protocol combining a spectral filter, background modeling, and subtraction.•Our methodology successfully extracts the CN stretching mode peak from methyl thiocyanate at remarkably low concentrations (0.25 mM) in water, previously a challenge for FTIR spectroscopy.This approach offers valuable insights and tools for more accurate FTIR measurements using weak vibrational probes. This enhanced precision can potentially enable new approaches to enhance our understanding of protein structure and dynamics in solution.

Keywords: Analysis of weak vibrational probes; Data processing; FTIR; Protein structure; Spectroscopy; Thiocyanate; Vibrational modes.

Graphical abstract

Fig. 1

Measurement of SCN peak for different MeSCN concentrations. (A) Raw detector intensity spectra shown with and without the implementation of the bandpass filter. In the orange spectrum wavelengths outside the bandpass filter window are blocked. (B) Spectra of MeSCN of concentrations 5, 10, 25, and 50 mM show thiocyanate peak sitting on top of the H₂O band. The MeSCN CN stretching signal appears in ∼2162 cm^-1 region of the spectrum. Within lower concentrations of MeSCN (0.25 mM, 0.5 mM and 1 mM) only the large background absorption of the water combination band mode is visible, and the SCN absorbance is too weak to observe without the background subtraction methods described here.

Fig. 2

Solvent subtraction leads to accurate lineshapes at low concentrations.  Spectra following solvent subtraction from samples with different MeSCN concentrations. Inset shows zoomed in frequency region for better visualization. All samples show a clear SCN peak centered at 2162 cm^-1.

Fig. 3

Background modeling with 2th to 4th-order polynomials. The raw measured spectra for the three lower concentration of MeSCN are shown together with their polynomial fits as indicated in each figure. Here the 3rd order polynomial was choosen to correct the absortion background of H₂O since it accurately models the baseline while avoiding overfitting.

Fig. 4

Comparison of polynomial fit subtraction and baseline substraction for SCN.  (A) Spectra following 3rd order polynomial subtraction. The polynomial subtraction produces a relatively clean SCN peak with relatively clean baseline correction. (B) Spectra after the subtraction solvent subtraction. Similar to the spectra shown in Fig. 2, the SCN peak is centered at 2162 cm^-1 for all the three samples with lowest MeSCN concentration. For better visualization of the peaks a Savitzky–Golay smoothing filter was used (black dashlines) as described in the text.