Technical Report
A Study of the Deisotope Method for Mass Spectra of Complex Synthetic Polymer Mixtures Consisting of Multiple Repeating Units
2024 Volume 13 Issue 1 Pages A0165
Details
2024 Volume 13 Issue 1 Pages A0165
In the interpretation step of mass spectra obtained from synthetic polymers, isotope peaks are typically intense and cannot be ignored, especially in the higher mass range. To reduce the complexity of the spectra, deisotope processing is used. In this study, a deisotope processing method that is effective even for mixtures of different types of polymers is investigated.
Mass spectrometry is a very powerful technique for the analysis of synthetic polymers,1,2) especially for the analysis of their detailed molecular structures, in addition to the average molecular weight determination.3) The high mass resolution is also a major advantage, as individual peaks are easily detected as discontinuous peaks in the spectrum, and information about individual components can be easily obtained, even if the sample is a complex mixture.4) However, each element that makes up synthetic polymers, such as carbon, hydrogen, and oxygen, has its own isotopes.5) This results in an isotopic distribution where the mass peak is observed in multiple peaks, which can complicate the mass spectrum or its analysis, even for a single molecule. In general, the higher the molecular weight, the greater the influence of isotopes. Therefore, the presence of isotopic peaks cannot be ignored, especially when analyzing synthetic polymer samples. To avoid this, deisotoping6,7) is used to extract only the monoisotopic peaks from the isotope distributions that are necessary for the analysis (Fig. 1). In the deisotope processing, only the monoisotopic peak is picked up and the signal intensities in the isotope distribution are integrated. If the isotope distributions overlap, they need to be processed separately. Deisotoping generally requires information on the elemental composition of the sample molecules detected in the mass spectrum. Some of the commercially available mass spectrum analysis software have their deisotope functions, and all of them require the elemental composition information of the monomer units (and optionally cation) as a parameter setting for deisotoping. However, as only one elemental composition can be entered, it may not be sufficiently helpful for a sample that is a mixture of several synthetic polymers with different monomer units. In many cases, mass spectra are analyzed by deisotoping followed by peak assignment, but in this study, a mass spectrum analysis method was investigated that can handle mixtures of synthetic polymers by deisotoping after peak assignment.
Bis(3-aminopropyl) terminated poly(dimethylsiloxane), (PDMS, average Mn ~2500, Sigma-Aldrich Co. LLC, St. Louis, MO, USA), polypropylene glycol (PPG, typical Mn: 2000 Da, Sigma Aldrich), and polyethylene glycol (PEG, typical Mn: 2000 Da, Sigma Aldrich) were obtained and mixed as tetrahydrofuran (THF) solution to be used as the samples. A 20 mg/mL THF solution of 2,5-dihydroxy benzoic acid (2,5-DHB, Bruker Corporation, Billerica, MA, USA) as the matrix-assisted laser desorption/ionization (MALDI) matrix was prepared. Also, 2 mg/ml THF solutions of sodium trifluoroacetate (NaTFA, Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) and silver trifluoroacetate (AgTFA, Sigma Aldrich) as cationizing agents were prepared. 0.5 µL of the sample-matrix-cationizing agent solution mixture was applied to the MTP groundsteel target plate (Bruker). MALDI-time of flight mass spectrometer (MALDI-TOF MS) instruments autoflex maX (Bruker) and ultrafleXtreme (Bruker) were used for mass spectrometry in positive reflector mode. The former was used for the analysis of the PPG/PDMS mixture sample, and the latter for the PEG/PPG mixture sample. Data processing and spectrum display of the mass spectra were performed using flexAnalysis 4.0.14 (Bruker) and Polymerix 3.01 (Sierra Analytics, Modesto, CA, USA). The flexAnalysis software is designed as a data viewer equipped with peak-picking functions. In this study, flexAnalysis was used to generate peak lists using its deisotoping function called SNAP (sophisticated numerical annotation procedure) algorithm. Polymerix is a synthetic polymer spectrum analysis software that has versatile functions, including peak picking, peak assignment, peak list export/import and so on. In this study, Polymerix was used for peak assignment, peak list export/import, and spectrum drawing.
Figure 2 shows the mass spectrum of the PPG/PDMS mixture sample. The spectrum is relatively complicated, mainly because the PDMS contains four components with different end-group structures (PDMS-1: C3H8N-(C2H6SiO)n-C5H14NSi, PDMS-2: C3H8N-(C2H6SiO)n-H, PDMS-3: HO-(C2H6SiO)n-H and PDMS-4: (C2H6SiO)n (cyclic structure without end-group)), in contrast to PPG which has only one structure (HO-(C3H6O)n-H). In addition, the peak in the inset of the figure is irregularly intense compared to other peaks due to the overlap of the PDMS-1 30 mer peak and the PPG 41 mer peak.
A combination of Polymerix and Excel 2016 (Microsoft, Redmond, WA, USA) was used for the deisotoping process in this study. The workflow tested in this study is summarized in Fig. 3, along with the other workflows for comparison. With regard to Step 1 in Workflows 1 and 3, peak picking in Polymerix was performed with the default settings except for the peak width setting (0.2 Da) without its deisotope function. In Workflow 2, the peak area information calculated using the SNAP algorithm is the sum of the peak areas of peaks in the isotopic distribution. Regarding Step 2 in Workflow 3, it is not necessary to find all the monoisotopic peaks in the spectrum, but it is enough to find two monoisotopic peaks from neighboring isotopic distributions that have a regular interval corresponding to the mass of the monomer unit. It would not be difficult to do this by searching for isotope distributions that do not overlap with the others. For Step 3 and subsequent steps in Workflow 3, since the peak intensity information from Polymerix is based on the peak area calculated from the profile spectrum, all intensity calculations in this study are based on the peak area or theoretical abundance ratio between isotopic species.
Figure 4A shows the results of peak assignment in Polymerix without deisotoping (Workflow 1). The peaks to be assigned are covered, but the peak intensity information is not appropriate as they are not deisotoped. Figures 4B and 4C shows the results of deisotoping using flexAnalysis software with PDMS and PPG, respectively as the monomer unit settings for the deisotope function (Workflow 2). With PDMS as the monomer unit setting in the flexAnalysis software, the series of peaks from PDMS are deisotoped and assigned relatively well, whereas the series of peaks from PPG are not assigned well. With PPG as the monomer unit setting the PPG peak series is well assigned, whereas the PDMS peak series are not. Figure 4D shows the deisotope results using the method investigated in this study (Workflow 3), which shows a good assignment of both PDMS and PPG. Even for a peak with a close m/z and overlapping isotopic distribution, the intensity is subtracted taking into account the theoretical isotopic distribution shape, so that irregular intensities due to the overlap of peaks around m/z 2500 are correctly assigned. In addition, unlike the case without the deisotoping process (Fig. 4A), where only the monoisotopic peaks are considered, the peak intensities reflect the entire isotopic distribution and more accurately reflect the molecular weight distribution of synthetic polymers. This is also beneficial for component composition analysis. However, there is still some scatter in the signal intensities which makes the polymer distribution less smooth. This is because the intensity calculation is based only on the intensity of a single monoisotopic peak and not on fitting the whole isotope distribution. This will be a future improvement. The complete results, including PDMS-3 and 4, are shown in the Supporting Information (Fig. S1).
Figure 5A shows the spectrum of the PEG/PPG mixture sample with a NaTFA/AgTFA mixture as the cationizing agent. In addition to being a PEG/PPG mixture, the detection of both Na+ and Ag+ adduct ions complicates the spectrum. Some of the peaks overlap, as shown in Fig. 5B, which is an example of three different ions, PEG 46 mer (Ag adduct), PEG 48 mer (Na adduct), and PPG 35 mer (Ag adduct) overlapping. Figure 6 compares the peak assignment results after each processing. Table 1 summarizes the peak assignment coverage and intensity ratio between the PEG-Na, PEG-Ag, PPG-Na and PPG-Ag signal series. Without deisotoping (Fig. 6A, Workflow 1), the peak assignment coverage (the ratio of correctly assigned peaks to the entire peak list obtained by the processing) shown in Table 1 is low at 20.41% because there are many non-monoisotopic peaks and their signal intensities are often higher than those of the monoisotopic peaks. Also, the ratio between the four polymer species/cation species combinations (PEG-Na:PEG-Ag:PPG-Na:PPG-Ag) is not appropriate because the peak intensities of the isotopic peaks within the isotopic distribution are not taken into account. Deisotoping using flexAnalysis with PEG-Na and PEG-Ag as monomer unit settings (Figs. 6B and 6C, Workflow 2) also results in insufficient peak assignment coverage (28.0% and 55.9% in Table 1) because there are many peaks that are wrongly recognized as monoisotopic peaks. In Table 1 the intensity ratios of PEG-Ag and PPG-Ag in deisotoping method B (0.5 and 1.1) are quite small. This is because the PEG-Ag signals could not be properly detected with the monomer/cation setting of PEG-Na. However, the ratio is not 0.0 because the flexAnalysis/SNAP algorithm works if the partial isotope pattern matches the theoretical isotope pattern, even if the setting is wrong. The same is observed for the intensity ratios of PEG-Na and PPG-Na in deisotoping method C (0.2 and 0.9). The deisotope method of this study works better even for the Ag-adduct peaks, which have their unique isotope distribution shapes (Fig. 6D, Workflow 3), and 100% coverage is obtained because only the peaks correctly identified as monoisotopic peaks in Step 2 of Workflow 3 are processed. The composition between the polymer peak series is also obtained appropriately, considering the peak intensities within the isotopic distribution, which is useful for compositional analysis applications. The ratios of PEG-Ag and PPG-Ag are larger in the result of Workflow 3 (Table 1D, 32.9 and 34.1, respectively) compared to the result of Workflow 1 (Table 1A, 25.9 and 25.2, respectively). This is because the intensities of monoisotopic peaks of Ag adduct ions tend to be low relative to other non-isotopic peaks compared to those of Na adduct ions. This is properly considered in the deisotope method of this study.
| Deisotoping method | Peak assignment coverage | Intensity ratio (PEG-Na:PEG-Ag: PPG-Na:PPG-Ag) |
|
| A | Workflow 1: without deisotoping |
20.4% | 39.1:25.9:9.9:25.2 |
| B | Workflow 2: Deisotoping by flexAnalysis (monomer/cation setting: PEG-Na) |
28.0% | 76.7:0.5:21.8:1.1 |
| C | Workflow 2: Deisotoping by flexAnalysis (monomer/cation setting: PEG-Ag) |
55.9% | 0.2:49.7:0.9:49.2 |
| D | Workflow 3: Deisotoping using the method of this study |
100% | 26.1:32.9: 6.9:34.1 |
Appropriate deisotoping would be required for proper characterization of polymer distribution and analysis of component composition. By working in a different order than usual, that is, deisotoping after peak assignment, it was possible to perform correct deisotoping for each of the multiple types of components in complex spectra acquired from mixed synthetic polymer samples. It was also possible to perform deisotoping well not only for multiple synthetic polymer mixtures but also for mixtures of multiple types of cation adducts. Further improvements may be required for the higher molecular weight range where monoisotopic peak intensities tend to be lower and the peak intensity calculation method needs to be made more accurate and robust. This is a subject for future work.
Figure S1. Peak assignment results of the PPG/PDMS mixture sample using Polymerix after (A) no deisotoping, (B) deisotoping using flexAnalysis with PDMS as the monomer unit setting, (C) deisotoping using flexAnalysis with PPG as the monomer unit setting, and (D) deisotoping using the method of this study. The top spectra show all components, the four middle spectra show PDMS components with different end groups (PDMS-1, 2, 3, and 4), and the bottom spectra show PPG components. Only correctly assigned peaks are shown.
Mass Spectrom (Tokyo) 2024; 13(1): A0165
