Molecular Characterization of High Molecular Weight Polyesters by Matrix-Assisted Laser Desorption/Ionization High-Resolution Time-of-Flight Mass Spectrometry Combined with On-plate Alkaline Degradation and Mass Defect Analysis
Matrix-assisted laser desorption ionization high-resolution time-of-flight mass spectrometry (MALDI HR TOF MS) is a powerful tool for the molecular characterization of industrial polymers. However, accurate mass determination and resolution of isobaric ions are possible for oligomer samples only typically below m/z 3000. To cut long polymer chains into oligomers suitable for high-resolution mass spectrometry, we propose a simple “on-plate” alkaline degradation of polyesters as a sample pretreatment technique prior to the MALDI TOF MS measurement. This pretreatment can be performed on a MALDI target using a small amount of sample (μg or less) and 1 μL of alkaline reagent by simple pipetting. Informative mass spectra in the oligomeric mass range are successfully recorded but complicated by the variation of end-groups and the copolymeric composition of the degradation products. Data processing is assisted by a series of advanced Kendrick mass defect (KMD) analyses recently proposed by the authors to plot visually understandable two-dimensional maps. On-plate degradation pretreatment, high-resolution MALDI TOF MS measurements, and advanced KMD analyses are innovatively combined for the compositional characterization of bacterial poly(3-hydroxybutyric acid-co-3-hydroxyvaleric acid) and industrial poly(ethylene terephthalate) samples.
KeywordsAlkaline degradation On-plate sample preparation High molecular weight polyesters MALDI TOF MS High-resolution mass spectrometry Kendrick mass defect analysis
Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI TOF MS) is an effective method for the characterization of polymers revealing their molecular weight distribution and the nature of their end-groups . Thanks to a remarkable progress in analytical instrumentation last decade, a cutting-edge mass spectrometer with a unique 17-m-long spiral-shaped TOF analyzer [2, 3] makes possible the discrimination of isobaric ions as a valuable feature for the characterization of polymers [4, 5, 6]. However, high-resolution/high-accuracy mass measurements are possible for oligomer samples only with a mass range no greater than m/z 3000. The oligomeric fraction (if existing) may nevertheless not reflect the whole polymer sample in terms of end-groups and copolymeric composition, casting doubts on the capability of MS for the characterization of industrial high molecular weight samples.
To overcome this issue, Arakawa and co-workers [7, 8, 9] introduced the ultrasonic degradation of polymers prior to their analysis by MALDI TOF MS. They reported that various end-groups were generated by sonication, while sufficient chain shortening required a long treatment time—typically several hours—due to low efficiency of the ultrasonic degradation. Owing to these limitations, the ultrasonic degradation has not been used in practice for the sample preparation so far.
Pyrolysis is more commonly used as a mean of generating low molecular weight products from synthetic polymers and natural products. Pyrolysis gas chromatography (Py-GC) is the preferred technique for the molecular and structural characterization of high molecular weight (as well as cross-linked) polymers . Flash pyrolysis at high temperature exceeding 400 °C is suitable to obtain monomeric products while information about the sequence in the original polymer is lost. Alternatively, partial pyrolysis at lower temperatures might be an effective approach to produce oligomeric products. In the case of polyesters, their pyrolysis is well known to form oligomers carrying carboxyl and olefin end-groups upon a β-hydrogen elimination. Based on this reaction, partial pyrolysis has been applied for the characterization of bacterial copolyesters in combination with fast atom bombardment (FAB)-MS , electrospray ionization (ESI)-MS , or MALDI TOF MS . In these reports, the sample was heated until it suffered a 20% weight loss as followed by thermogravimetric (TG) measurement. As such, it required a few milligrams of sample for accuracy reasons. Reproducing the experiments, we found that the temperature control is difficult and that a dedicated rapid cooling apparatus is required to immediately quench the degradation reaction at 20 wt% weight loss.
Acid- or base-catalyzed transesterification would be another approach to softly decompose high molecular weight polyesters into oligomers. Prior to MALDI TOF MS measurements, hydrochloride and sodium methoxide in methanol were used as the reagents for the partial transesterification of poly(3-hydroxybutyric acid-co-3-hydroxyvaleric acid) (P(3HB-co-3HV))  and poly(β-hydroxyalkanoate)s (PHAs) , respectively. However, these reactions were carried out at flask scale. It may thus require a trial-and-error optimization of the reaction conditions to produce the degradation products having suitable molecular weights for a high-resolution mass spectral characterization.
In this paper, we propose a simple pretreatment method for MALDI TOF MS measurements of high molecular weight polyesters as a modification of the partial alkaline transesterification previously reported . In essence, the main feature of our method is that the pretreatment of a polymer sample—either a thin layer drop-casted from a solution or a film—can be performed on a MALDI target plate within a few minutes by pipetting an alkaline solution. This idea originated from the industrial processing of polyester fibers. In the field of textile processing, surface modification by alkaline treatments has been used to improve the properties of polyester fibers as a finishing technique . In tissue engineering, the cell affinity of polyester-based materials has also been improved via their surface modification by alkaline hydrolysis treatment [17, 18, 19, 20]. In both cases, molecular weight reduction of polyesters has been found to occur at the surface layer only. This phenomenon would be convenient for the characterization of high molecular weight polyesters by MALDI TOF MS. Despite a wide range of molecular weights from oligomeric degradation products to intact polymer chains, the surface layer of a partially degraded sample would indeed concentrate the oligomers without any optimization of the reaction conditions.
A last difficulty remains considering that the scission of ester bond by transesterification produces several different types of end-groups. For the characterization of such complex blends, the authors have proposed a new strategy combining the high-resolution MALDI TOF MS measurements with a Kendrick mass defect (KMD) analysis [6, 21]. A KMD analysis  is a data processing tool for the rapid visualization of complex mass spectra traditionally used for petroleomics [23, 24] but recently extended to the field of polymer analysis [6, 21, 25, 26, 27, 28]. The two-dimensional plot generated by a KMD analysis displays a relationship between the molecular weight (or degree of polymerization; x-axis) and differences in elemental composition reflecting differences in chemical structures (variation of end-groups, comonomer distribution, degree of modification, oxidation, etc.; y-axis). More recently, our research group has aggressively proposed a series of advanced KMD analyses for the compositional characterization of copolymers by introducing, refining, or extending the concepts of “fractional base unit” [29, 30, 31] and “referenced KMD” [32, 33].
In this paper, a simple alkaline degradation on the MALDI target plate (termed “on-plate degradation”) is first investigated through the MALDI TOF MS analysis of a commercial polyester. Isotopic labeling (reagents and solvent) is used to confirm the degradation scheme. The compositional characterization of high molecular weight bacterially produced copolyester and industrial polyester is then carried out to demonstrate that combining the on-plate degradation and the advanced KMD analysis pushes the boundaries of high-resolution MALDI TOF MS.
Chemicals and Materials
As an alkaline reagent for the on-plate degradation, sodium hydroxide (NaOH, reagent grade, Wako pure chemicals, Osaka, Japan) was dissolved in methanol (CH3OH, infinity pure grade, Wako) or methanol-d4 (99.8 atom%, Wako) at ca. 10 mg mL−1. A 28% solution of sodium methoxide (CH3ONa) in methanol (Wako) was diluted to ca. 10 mg mL−1 in methanol. All deuterated alkaline reagents of sodium deuteroxide (NaOD) in methanol-d4 were prepared as follows: ca. 10 μL of a 40% solution of NaOD in deuterium oxide (D2O, 99.5 atom%) purchased from Alfa Aesar (Lancashire, UK) was vacuum-dried to remove D2O and ca. 40 μL of methanol-d4 was added to form a methanolic solution at ca. 10 mg mL−1. Distilled water or D2O (100 atom%, Wako) was used to wash the excess amount of the alkaline reagent on the MALDI target.
Tetrahydrofuran (THF, reagent grade), 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP), and chloroform (HPLC grade) were from Wako and used as solvent for the polymer samples and/or as mobile phase for size-exclusion chromatography (SEC). 2,4,6-Trihydroxyacetophenone (THAP, ultrapure grade, Protea Biosciences, West Virginia, USA) was used as a MALDI matrix and dissolved in THF at ca. 20 mg mL−1.
SEC Measurement and Fractionation
As a fundamental study, the SEC fractionation was carried out to confirm the degradation products of high molecular weight polyesters. Average molecular weights of P3HB and P(3HB-co-3HV) were determined by a build-up LC system (Tosoh, Tokyo, Japan) composed of an AS-8020 auto sampler, a CCPS pump, an SD-8022 degasser, a CO8020 column oven, and a RI-8020 refractive index detector (RID). Ten microliters of the sample solution at 2 mg mL−1 in HFIP were injected and analyzed using two TSKgel super HM-H columns (6.0 mm × 150 mm) connected in series following an HM-H guard column with HFIP containing 0.5 mM sodium trifluoroacetate as the mobile phase (0.3 mL/min). A calibration curve was made using a kit of poly(methyl methacrylate) standards (PMMA M-75 kit, Shodex, Tokyo, Japan).
A high molecular weight fraction of PCL was recovered from a SEC elution of the commercial PCL using a HLC8220 system (Tosoh) equipped with a RID. The sample solution (200 μL at 2 mg mL−1 in CHCl3) were fractionated using two TSKgel multipore HXL-M columns (7.8 mm × 300 mm) connected in series following a multipore Hxl guard column with CHCl3 as the mobile phase (1 mL min−1). Aliquots of 0.5 mL (30 s of elution) were collected in vials directly after the RID. The concentration of the chosen fraction could be estimated as 0.08 mg mL−1 based on the RID signal intensity. The number-average molecular weight (Mn) of the fractionated sample was about 13,800 g mol−1 (Supporting Information Figure SI-1, MALDI TOF mass spectrum in linear mode) and no peaks were observed in the MALDI spiral TOF mass spectrum under m/z 10,000.
On-plate Degradation Procedure
All samples but PETs were dissolved in THF at ca. 1 mg mL−1. About 1 μL of each sample solution was drop-casted on a disposable MALDI target plate (non-focus, non-hydrophobic grade, Hudson Surface Technology, New Jersey, USA) to form a polymer thin layer. The PET samples were cut by a micro puncher (inner diameter 0.75 mm, Frontier Lab., Koriyama, Japan) and directly put on the MALDI target plate. Next, ca. 1 μL of HFIP was added on each PET pellet to form a polymer thin layer around it. Then, ca. 1 μL of the alkaline reagent was deposited over the sample layer and left to air-dry for about 5 min until methanol fully evaporated. The excess amount of the alkaline reagent was washed with ca. 5 μL of distilled water or D2O by pipetting (discharge/suction repeated three times) following the same procedure as an on-plate desalting for MALDI sample preparation . After drying, ca. 1 μL of the matrix solution was applied on the sample spot.
MALDI TOF MS Measurements and Data Processing
MALDI TOF MS measurements were performed using a JMS-S3000 spiralTOF™ (JEOL, Akishima, Japan). This apparatus is equipped with an Nd: YFL laser (λ = 349 nm) and a spiralTOF™ analyzer with a spiral ion trajectory corresponding to a flight length of approximately 17 m. The apparatus parameters were set to maintain ∆M < ca. 0.03 Da at FWHM over the range of m/z 800–3000. Mass calibration was made with a PMMA standard (peak-top molecular weight, Mp = 1310) purchased from Polymer Laboratories (Shropshire, UK).
Mass spectra were processed and exported using MS Tornado Analysis software (JEOL). KMD analyses of the exported MS data were performed using msRepeatFinder software ver. 2.1.0 (JEOL). The referenced KMD analysis to depict the compositional distribution of P(3HB-co-3HV) was eventually carried out using Excel and a home-made set of macros (VBA coding).
Results and Discussion
Fundamental Study of the On-plate Degradation
From these results, we can infer that the origin of hydrogen atoms in the carboxyl and hydroxyl end-groups would be the PCL molecules themselves. The alkaline degradation of polyesters has been well investigated in the literature. The methoxide anion in the reaction solvent nucleophilically attacks the carbonyl of the ester bond producing methyl esters. If the alkaline degradation using deuterium-labeled reagents is performed completely to yield monomeric products, the hydrogen atoms of their hydroxyl groups should all be replaced with deuterium. However, the alkaline degradation reaction by the on-plate degradation procedure is surely not complete as it occurs at the surface layer only, leaving a considerable amount of intact carboxyl end-groups near the reaction sites. So, the protons released from the carboxyl end-groups would connect with alkoxide anion and form hydroxyl end-groups free of deuterium despite using deuterium-labeled reagents.
Application for the Compositional Analysis of High Molecular Weight Copolymers
Amid growing interest in environmental issues, PHAs produced by bacteria are seen as one of the most promising candidates for renewable materials. Since the chemical information such as the type of monomers and the copolymeric composition of PHAs generally reflects the carbon source fed to the bioreactor , the analysis of PHAs is critical from the view point of quality control. However, the high molecular weight of bacterial PHAs typically over hundreds of thousands Da makes the mass analysis of intact PHAs difficult especially for the evaluation of co-monomeric content . In this study, the capability of the on-plate degradation combined with high-resolution MALDI TOF MS was demonstrated for the thorough characterization of high molecular weight PHAs.
The on-plate degradation as the pretreatment for the high-resolution MALDI TOF MS measurements was further applied for the compositional characterization of P(3HB-co-3HV) copolymer. An isomeric issue arises here as the mass difference of 14 Da (CH2) between 3HB and 3HV co-monomeric units is the same as the mass variation caused by the esterification of the carboxyl group. To overcome this issue, the deuterium-labeled methanol-d4 was used as the degradation solvent to induce a mass variation + 17 Da (+CD3 –H, + 3 Da shift) for the methyl esters as described in the previous section.
The compositional characterization of copolymers requires a full assignment for a huge number of observed peaks which remains a tedious task. Even if an automatic peak picking using appropriate programs would help for this purpose, a perfect mass calibration is still needed over the whole mass range. Otherwise, inappropriate single setting for the mass tolerance often leads to incorrect peak picking especially for complicated mass spectra. In order to facilitate the data processing avoiding the risk of the incorrect peak assignment, we have proposed a series of advanced KMD analysis techniques [29, 30, 31, 32, 33]. The final goal of this section is to display a degree of polymerization (DP) plot  as an instant graphical visualization of the comonomer distribution of P(3HB-co-3HV).
Usually, data triplets (NKMr/X, KMDr/X, abundance) are plotted in a two-dimensional “bubble chart” with NKMr/X on the x-axis and KMDr/X on the y-axis to produce a KMD plot. However, in this report, nominal mass of m/z was set for the x-axis, because NKMr/X greatly deviates from nominal mass depending on the value of X.
The next step was the rearrangement of the extracted dataset by another resolution-enhanced KMD analysis with an alternative divisor to better expand the comonomer distribution (detailed data processing is described in the Supporting Information). The resolution-enhanced KMD plot for the extracted type I is shown in Figure 6c, where X = 62 was set for 3HB. The components of type I are now mainly separated based on their content in 3HV (one horizontal line per 3HV unit) while points within a given line differ by their content in 3HB.
The DP plot is finally plotted by displaying the triplets (DP(3HB), DP(3HV), abundance) in a two-dimensional bubble chart (Figure 6d). Detailed explanation to produce the DP plot is additionally described in the Supporting Information. It graphically displays the comonomer distribution of the degraded P(3HB-co-3HV). The most abundant products are P3HB homopolymeric oligomers (DP(3HV) = 0) while no P3HV homopolymeric oligomers are observed (DP(3HB)min = 5), suggesting that the original high molecular weight copolymer is a random copolymer with high 3HB content. A block architecture would have indeed released P3HV homopolymeric oligomers upon on-plate degradation. Based on the DP plot, the average 3HV content is calculated at 8.9 ± 0.2 mol% for three sample spots in moderate agreement with the 12 mol% content provided by the supplier. Such a rapid compositional characterization of a high molecular weight copolyester out of reach of a direct MALDI TOF MS analysis has been made possible by the successful combination of the on-plate degradation and the advanced KMD analysis.
Application for the Characterization of PET Film and Bottle Samples
Considering that the partial pyrolysis preparation requires a few mg of sample for the TG pyrolysis [11, 12, 13], there is a clear advantage using the on-plate degradation technique which requires ten times lower amount of sample. Accordingly, the next application was proposed with a view to local analysis of industrial materials. Industrial PET film and bottle were subjected for the on-plate degradation as model samples with ca. 14 μg of PET and 1 μL of NaOH methanol solution.
The on-plate degradation procedure based on a partial alkaline transesterification can easily yield oligomeric products from high molecular weight polyesters. This simple and minute-long pretreatment can be performed directly on the MALDI target by pipetting. Since small amounts of sample and reagents are required, isotope labeling is readily done to shift the mass of specific products and track some degradation pathways or avoid isomeric issues. The resulting complex mass spectra with products having different end-group combinations or copolymer compositions—informative but uneasy to decipher manually—were efficiently processed using a series of advanced KMD analysis techniques. A resolution-enhanced KMD plot, a remainder plot, and a DP plot helped at visualizing the molecular weight distribution of oligomers having specific end-group combinations, the distribution of these end-group combinations, and the comonomer distribution. As demonstrated with the characterization of bacterial polyesters and industrial PET samples, the combination of the on-plate degradation pretreatment, high-resolution MALDI TOF MS measurements, and advanced KMD analyses constitutes an effective analytical strategy for the molecular characterization of high molecular weight polyesters.
The authors thank Akihiro Oishi of the National Institute of Advanced Industrial Science and Technology (AIST) for his support of SEC measurements.
- 8.Watanabe, T., Okabayashi, M., Kurokawa, D., Nishimoto, Y., Ozawa, T., Kawasaki, H., Arakawa, R.: Determination of primary bond scissions by mass spectrometric analysis of ultrasonic degradation products of poly(ethylene oxide-block-propylene oxide) copolymers. J. Mass Spectrom. 45, 799–805 (2010)CrossRefGoogle Scholar
- 12.Barbuzzi, T., Giuffrida, M., Impallomeni, G., Camazza, S., Ferreri, A., Guglielmino, S.P.P., Ballistreri, A.: Microbial synthesis of poly(3-hydroxyalkanoatels) by Pseudomonas aeruginosa from fatty acids: identification of higher monomer units and structural characterization. Biomacromolecules. 5, 2469–2478 (2004)CrossRefGoogle Scholar
- 13.Impallomeni, G., Ballistreri, A., Carnemolla, G.M., Franco, D., Guglielmino, S.P.P.: Matrix-assisted laser desorption/ionization time-of-flight vs. fast-atom bombardment and electrospray ionization mass spectrometry in the structural characterization of bacterial poly(3-hydroxyalkanoates). Rapid Commun. Mass Spectrom. 29, 811–820 (2015)CrossRefGoogle Scholar
- 14.Abate, R., Ballistreri, A., Montaudo, G., Garozzo, D., Impallomeni, G., Critchley, G., Tanaka, K.: Quantitative applications of matrix-assisted laser-desorption ionization with time-of-flight mass-spectrometry - determination of copolymer composition in bacterial copolyesters. Rapid Commun. Mass Spectrom. 7, 1033–1036 (1993)CrossRefGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.