Molecular Formula Prediction for Chemical Filtering of 3D OrbiSIMS Datasets

Abstract

Modern mass spectrometry techniques produce a wealth of spectral data, and although this is an advantage in terms of the richness of the information available, the volume and complexity of data can prevent a thorough interpretation to reach useful conclusions. Application of molecular formula prediction (MFP) to produce annotated lists of ions that have been filtered by their elemental composition and considering structural double bond equivalence are widely used on high resolving power mass spectrometry datasets. However, this has not been applied to secondary ion mass spectrometry data. Here, we apply this data interpretation approach to 3D OrbiSIMS datasets, testing it for a series of increasingly complex samples. In an organic on inorganic sample, we successfully annotated the organic contaminant overlayer separately from the substrate. In a more challenging purely organic human serum sample we filtered out both proteins and lipids based on elemental compositions, 226 different lipids were identified and validated using existing databases, and we assigned amino acid sequences of abundant serum proteins including albumin, fibronectin, and transferrin. Finally, we tested the approach on depth profile data from layered carbonaceous engine deposits and annotated previously unidentified lubricating oil species. Application of an unsupervised machine learning method on filtered ions after performing MFP from this sample uniquely separated depth profiles of species, which were not observed when performing the method on the entire dataset. Overall, the chemical filtering approach using MFP has great potential in enabling full interpretation of complex 3D OrbiSIMS datasets from a plethora of material types.

Figure 1

Separating organic and inorganic species on aluminum foil 3D OrbiSIMS depth profiling datasets using MFP. (a) Optical image of the analysis area before Ar₃₀₀₀⁺ depth profiling. (b) Raw depth profile accumulation spectra in the negative ion mode. (c) Positive scale, plots of the inorganic species containing Al, O, and H identified using MFP. Negative scale, contaminants identified after performing a repeated mass separation of the initial dataset, followed by another formula filtering iteration analysis to identify species with C, H, O, N, S, and P (intensity × −1). (d) Depth profiles of some species in each sub-group. MFP, molecular formula prediction.

Figure 2

(a) Comparison of all assignments proposed by MFP in a collated peak list of all manually assigned protein peaks in 16 different protein samples (black) and only correct protein assignments (red). (b) All protein peaks assigned in 16 protein samples are widely spread around the trend line of DBE/C_n (grey). The variability of DBE/C_n of protein assignments is caused by different amino acid DBE/C_n (orange—asparagine, R; yellow—leucine, L; red—histidine, H, green—asparagine, N; brown—phenylalanine, F). (c) Human serum sample is too complex to manually assign fragments of biological molecules. (d) MFP combined with the LIPID MAPS database automatically assigns lipid groups (black). Additional MFP on the rest of the spectrum enables assignment of proteins in serum (red). Peaks assigned as fragments of fibronectin (C₁₅H₂₂N₄O₃Na⁺, m/z 329.1585), transferrin (C₂₄H₄₄N₆O₅Na⁺, m/z 519.3258), and human serum albumin (C₂₄H₄₀N₈O₇Na⁺, m/z 575.2918). Chemical filtering was carried out on SIMS-MFP software.

Figure 3

Using MFP to search for sulfated species in gasoline engine deposits. In each case, the raw 3D OrbiSIMS depth profile accumulation spectrum is shown above the spectrum of sulfated ions and the lowest is the filtered DBE plot of these species. (a) Deposit on injector tip 1, showing benzyl sulfonates present up to high masses (C₇₅). (b) Injector tip deposit 2, showing benzyl sulfonates <C₄₀ and a unique distribution of alkyl sulfonates with a DBE value <2. (c) Results from the injector needle, showing only benzyl sulfonates but to a lower maximum mass than other samples.

Figure 4

Demonstrating MVA on datasets after performing MFP on 3D OrbiSIMS depth profiling data from injector tip 1 deposit. (a) DBE versus carbon number plots of all oxygenated ions (C_nH_nO₂^–) identified using MFP. (b) DBE versus carbon number plots of all carbonaceous ions (C_nH_<1^–) identified using MFP. (c) Loadings of oxygenated ions after performing NMF on the filtered list of oxygenated ions identified using MFP. (d) The first two loadings of carbonaceous ions identified using MFP. (e) Depth profiles of the first two endmembers of the oxygenated ions. (f) Depth profiles of the first two endmembers of the carbonaceous ions.