Mass Spectrometry Imaging of Time-Dependently Photodegraded Light Stabilizers in Polyethylene Films Using Tapping-Mode Scanning Probe Electrospray Ionization

Abstract

Light stabilizers are additives that are widely used to improve the lifespan and performance of polymer materials. To develop advanced polymer materials, analytical techniques investigate the degradation mechanisms and distribution of additives in polymers are crucial. Herein, two extraction–ionization methods were used: tapping-mode scanning probe electrospray ionization (t-SPESI) and liquid extraction surface analysis (LESA). The distribution and molecular structure of the photodegradation products were investigated using polyethylene films containing two types of oligomeric hindered amine light stabilizers (o-HALS). In addition, to study the relationship between light irradiation time and the relative amount of photodegradation products, we developed a method for preparing films with multiple photodegradation regions. Mass spectrometry imaging (MSI) using t-SPESI (t-SPESI-MSI) revealed that the signal intensities of HALS decreased with the time of light irradiation, and its degradation products progressively changed. Moreover, tandem mass spectrometry (MS/MS) using LESA (LESA-MS/MS) revealed that degradation products were generated by HALS fragmentation in the polymer film. By integrating these results, we propose multiple and stepwise reactions for the formation of the photodegradation products. Results indicate that the combined use of t-SPESI-MSI and LESA-MS/MS can directly analyze and understand the photodegradation mechanism of o-HALS in polymer materials.

INTRODUCTION

Polymer materials are crucial components in a wide range of industries such as semiconductors, batteries, and medical fields. The properties and shapes of polymer materials are controlled according to their intended use; however, exposure to high temperatures or sunlight can cause oxidative reactions to progress, leading to polymer degradation. Irradiation with high-energy ultraviolet light specifically introduces radicals and hydroxyl radicals into polymers, initiating oxidation and main chain scission reactions,¹⁾ leading to a decrease in the strength and discoloration of polymer materials. To protect polymers from light-induced degradation, hindered amine light stabilizers (HALS) have been used. Photooxidation transforms the amine group in the tetramethylpiperidyl component into a nitroxyl radical. This radical then interacts with polymer-derived radicals to form an amino ether intermediate. Subsequently, this intermediate engages with another radical species from the polymer, leading to the regeneration of the nitroxyl radical while simultaneously eliminating nonradical byproducts. This reversible chemical reaction, known as the Denisov cycle,^2–6) inhibits radical chain reactions that result in polymer photodegradation. Polymers formulated with HALS compounds can still degrade under light exposure.^7,8) However, research conducted on the photodegradation reactions of the tetramethylpiperidyl component and other chemical structures of HALS is limited. Hence, the development of analytical methods to examine HALS and their degradation products within polymers is crucial to improve product quality.

HALS can be classified into low molecular-weight monomeric HALS (m-HALS) and oligomeric HALS (o-HALS) with a molecular weight distribution of up to several thousands.⁹⁾ m-HALS has been widely used owing to their high dispersibility in polymers. Incorporating m-HALS into polymers with low crystallinity, such as polyethylene, leads to surface migration over time. This phenomenon may result in volatilization and sublimation, potentially compromising the photodegradation resistance of the material.^10,11) By contrast, o-HALS are distinguished by its high molecular weight, which suppresses diffusion over time.

Typically, m-HALS and o-HALS analyses via mass spectrometry (MS) involve a combination of solvent extraction and chromatography. The established analysis method for m-HALS is straightforward, as they can be extracted with a wide variety of solvents.¹²⁾ In comparison, o-HALS exhibit lower solubility in solvents than m-HALS. Consequently, suitable pretreatment methods, such as solvent extraction or dissolution precipitation, must be used to isolate and analyze o-HALS from polymers.^13,14) To date, o-HALS analysis has been conducted using high-performance liquid chromatography (HPLC)-electrospray ionization-mass spectrometry (ESI-MS), HPLC-atmospheric pressure chemical ionization-MS (APCI-MS),^15,16) matrix-assisted laser desorption/ionization-MS (MALDI-MS),^17,18) and pyrolysis gas chromatography/MS (GC/MS).^19,20) These methods have been primarily used to analyze the solvent extracts of samples; hence, the spatial information of photodegraded o-HALS in polymer materials is lost. When HALS dispersibility is low owing to the composition of the polymer or molding conditions, local degradation due to light can cause several issues such as bleed-out, poor adhesion, and discoloration. Therefore, a method to directly analyze o-HALS in the local regions of polymer materials using MS is expected to provide new information on photodegradation.

Direct analytical methods, such as direct analysis in real time (DART),²¹⁾ desorption electrospray ionization (DESI),^3,22,23) liquid extraction surface analysis (LESA),^24,25) and tapping-mode scanning probe electrospray ionization (t-SPESI),²⁶⁾ have been used to analyze polymer materials. Detection of m-HALS using DART²⁷⁾ and that of m-HALS and its degradation products using DESI and LESA have been reported.^3,25) These methods are characterized by minimal sample pretreatment or no pretreatment and the ability to directly and rapidly detect analytes from microregions. Specifically, DESI, LESA, and t-SPESI use solvents to extract and ionize components from the microregions of the sample. By combining LESA with tandem mass spectrometry (MS/MS) with collision-induced dissociation (CID), we can obtain a product ion spectrum to estimate the structure of the components in the local regions.²⁵⁾ Mass spectrometry imaging (MSI) can be performed by moving the extraction–ionization area across the sample surface, allowing visualization of the distribution of different components within the sample. In addition, MSI of m-HALS in coil coatings using DESI has been reported,³⁾ but not with LESA or DART as far as we could find. In DESI, charged droplets produced by electrospray are accelerated by a stream of nitrogen gas and directed onto the sample surface, facilitating the desorption and ionization of its components. The electrostatic repulsion between charged droplets widens the extraction region, which typically leads to a spatial resolution ranging from tens of micrometers or larger. In t-SPESI, the MSI pixel size can be reduced to 5 μm using an oscillating capillary and low-flow solvent.^28,29) Meanwhile, LESA uses a micropipette for solvent extraction and ionization, resulting in a spatial resolution of ~1 mm. MSI of the micron region using LESA is difficult, and although there are reports of analyzing m-HALS, there are no reports of direct analysis and MSI of o-HALS.

Herein, we report the photodegradation products of o-HALS in polyethylene films using t-SPESI-MSI and LESA-MS/MS. Two types of additives (Chimassorb 944 and Tinuvin 622) were used for o-HALS (Fig. 1). To investigate the relationship between light irradiation time and the relative amount of photodegradation products, we prepared a sample with multiple regions of different light irradiation times within a single film (Fig. 2). In addition, the observation area was set to include multiple light irradiation areas, and t-SPESI-MSI was performed. Photodegradation products whose distribution changed depending on the light irradiation time were selected, and the candidate ions of the photodegradation products were determined. Moreover, LESA-MS/MS was applied to these ions, and the structures of the photodegradation products were estimated.

Fig. 1. Chemical structure of (A) Chimassorb 944 and (B) Tinuvin 622.

Fig. 2. Overview of light irradiation tests and analysis.

EXPERIMENTAL

Chemicals

Chimassorb 944, poly [[6-[(1,1,3,3-tetramethylbutyl)amino]-1,3,5-triazine-2,4-diyl][(2,2,6,6-tetramethyl-4-piperidinyl)imino]-1,6-hexanediyl[(2,2,6,6-tetramethyl-4-piperidinyl)imino]] (BASF Japan Ltd., Tokyo, Japan), and Tinuvin 622, poly (4-hydroxy-2,2,6,6-tetramethyl-1-piperidineethanol-alt-1,4-butanedioic acid) (BASF Japan Ltd., Tokyo, Japan), were used as o-HALS. Figure 1 presents their chemical structures. In addition, a 100-μm-thick low-density polyethylene film containing 1 wt% each of Chimassorb 944 and Tinuvin 622 (Kyodo Chemical Company Limited, Hadano, Japan) was used. Polyethylene pellets containing 1 wt% were prepared by dry blending the masterbatch pellets (20 wt%) with raw polyethylene pellets in a single screw extruder at 180°C. Film samples were then obtained by film forming at 160°C. Moreover, N,N-dimethylformamide (DMF), methanol (MeOH), isopropyl alcohol (IPA), tetrahydrofuran (THF), and their mixtures were used as the extraction solvents. All solvents were of HPLC grade and purchased from Nacalai Tesque (Kyoto, Japan).

Fabrication of polymer films containing o-HALS with controlled photodegradation regions

Figure 2 presents the sample preparation procedure. A polyethylene film (2 × 3 cm) was partially wrapped with a 25-μm-thick aluminum foil. The foil was secured by wedging it between two bar-shaped neodymium magnets that were aligned parallel to the foil-covered region. The edge of the aluminum foil outside the MSI region was marked with a black marker. The film with aluminum foil was irradiated using a very high-pressure mercury vapor lamp (HB-10201AA, Ushio Inc., Tokyo, Japan) with absorption in the UV-A region (300–420 nm) and visible region (500–600 nm). The light intensity was set to 400 W/m² at 365 nm, without a bandpass filter. The position of the aluminum foil was shifted by changing the light irradiation time (0, 5, 20, 60, 180, and 360 min).

LESA-MS and LESA-MS/MS

LESA-MS and LESA-MS/MS were performed using a TriVersa NanoMate robotic system (TriVersa NanoMate LESAplus, Advion Inc., Ithaca, NY, USA) coupled to an Orbitrap mass spectrometer (Q Exactive Plus, Thermo Fisher Scientific Inc., Waltham, MA, USA). The extraction solvent (2 μL) was manipulated with a robotic pipette tip to form a liquid microjunction with a diameter of ~2 mm between the tip and the sample. The contact time between the solvent and the sample surface was 1 s, with 3 repetitions of aspiration and dispensation of the solvent. In addition, the droplets were electrosprayed through a 5-μm-diameter nozzle. LESA-MS experimental conditions were a spray voltage of 1.8 kV and nitrogen gas pressure of 0.3 psi. Moreover, unirradiated polyethylene films were measured by varying the composition ratio of the solvent mixtures (DMF/MeOH = 0/100, 50/50, 100/0 [v/v]; IPA/THF = 0/100, 25/75, 50/50, 75/25, 100/0 [v/v]) to optimize the extraction solvent. The signal intensities of the ions derived from Chimassorb 944 and Tinuvin 622 were investigated.

LESA-MS/MS of the polyethylene films was performed on the surface at 0, 20, 60, and 360 min of light irradiation. Product ion scans were performed for precursor ions determined using t-SPESI-MSI (m/z 601.563, 461.396, 629.558, 363.322, 417.370, and 403.354) for structural identification. The normalized collision energy³⁰⁾ was set between 15 and 55. Mass calibration was carried out by employing ESI with Pierce LTQ Velos ESI Positive Ion Calibration Solution (Thermo Fisher Scientific Inc., Waltham, MA, USA) containing caffeine, MRFA, and Ultramark 1621.

t-SPESI-MSI

Herein, a t-SPESI measurement system connected to a quadrupole-time-of-flight mass spectrometer (LCMS-9030, Shimadzu Corporation, Kyoto, Japan)²⁸⁾ was used. Silica emitters (i.d. 10 μm; Fossiliontech, Madrid, Spain) were used as capillary probes. The solvent was supplied at a flow rate of 10 nL/min using a nanoflow pump (LC-20A nano, Shimadzu Corporation) through a 5-m silica capillary toward the probe. The oscillation frequency of the probe was 664 Hz.

Line analysis (8-mm probe scanning) of the unirradiated polyethylene film was performed using two types of mixed solvents (DMF/MeOH = 50/50 [v/v] and IPA/THF = 75/25 [v/v]) to compare the ion signal intensities. Mass spectra were acquired in positive ionization mode in the m/z range of 200–3000. The scan speed of the probe on the surface was 400 μm/s. Mass spectra were measured during the 100 ms period (distance: 40 μm) of the probe scan. Each mass spectrum obtained was considered a different pixel to generate an ion image. The voltage applied to the solvent through the stainless-steel union was +3.4 kV. The ion transfer tube was connected to the desolvation line of the mass spectrometer and heated at 250°C. The areas of the irradiated polyethylene films used for MSI were 8 and 3 mm in the horizontal and vertical directions, respectively, and the pixel size was 40 μm. Other experimental parameters were the same as those used in the line analysis. The other measurement scheme was the same as that previously reported.²⁸⁾

Multiple data files acquired for all probe scans were converted into a single imaging data file using an in-house data converter. The imaging data were analyzed using a commercial software (IMAGEREVERAL, Shimadzu Corporation, Kyoto, Japan). Each area with different light irradiation times was set as the region of interest (ROI). To examine the ion signal intensities that varied after light irradiation, the average ion signal intensities and their standard deviations of all mass spectra contained in each ROI were calculated. The boundaries for each light-irradiated region were determined based on the marks on the film, images of the total ion current, and ion images of the degradation products of Chimassorb 944. Furthermore, partial least squares (PLS) regression was applied to the averaged ion signal intensities of the ROIs to identify ions whose signal intensities changed depending on the duration of light irradiation. The highest relative ion intensities detected at 20, 60, and 360 min of irradiation were set as predefined classes, and the PLS regression coefficients for each detected ion were calculated.

RESULTS AND DISCUSSION

Investigation of solvents for o-HALS extraction and ionization

To analyze o-HALS via t-SPESI and LESA, an appropriate solvent for extraction and ionization is required. Therefore, we first compared the signal intensities of the ions derived from Chimassorb 944 and Tinuvin 622 to determine the solvent for each ionization method. We examined unprocessed o-HALS-containing polyethylene films using two types of solvent mixtures: DMF/MeOH and IPA/THF. The solvent mixture of DMF/MeOH = 50/50 (v/v) was previously reported to be suitable for phospholipid extraction and ionization in biological tissues via t-SPESI.³¹⁾ Phospholipids are amphiphilic molecules with hydrophilic heads and hydrophobic acyl chains, indicating their potential applicability in o-HALS extraction. The less-polar solvent mixture of IPA/THF was expected to efficiently extract o-HALS from the polymers.

Figure 3A and 3B present the signal intensities of the protonated molecules derived from Chimassorb 944 and Tinuvin 622 obtained via LESA-MS. Higher ion signal intensities were obtained in IPA/THF than in DMF/MeOH. Figure 3C and 3D present the mass spectra obtained via LESA-MS. The ion peaks in the mass spectra are listed in Table 1. When IPA/THF was used, the ions of Chimassorb 944 and Tinuvin 622 were detected. However, when DMF/MeOH was used, the ion signal intensities of Chimassorb 944 decreased, and those of Tinuvin 622 were below the detection limit. Based on these comparisons, IPA/THF = 75/25 was used for LESA-MS/MS to investigate the molecular structure of photodegraded o-HALS. Chimassorb 944 and Tinuvin 622 compounds have molecular weights of 500–3000 and 300–4000, respectively. In this experiment, ions in the low molecular-weight region were measured. Conceivable reasons for the failure to detect high molecular weight ions are low extraction efficiency and ion suppression of high molecular weight components due to preferential ionization of low molecular weight components. It is likely that o-HALS interacts with the polymer chains and is stabilized. In particular, high molecular-weight o-HALS may not have been sufficiently extracted after a short extraction time in LESA.

Fig. 3. Comparison of the sum of the ion signal intensities derived from (A) Chimassorb 944 (m/z 993.951, 497.480, and 796.750, as shown in Table 1) and (B) Tinuvin 622 (m/z 316.212, 599.390, and 882.569, as shown in Table 1) in unirradiated polyethylene film by LESA-MS. The error bars represent the standard deviation of 3 repeated measurements. Comparison of the mass spectra obtained by LESA using (C) DMF/MeOH = 50/50 and (D) IPA/THF = 75/25, and the mass spectra obtained by t-SPESI using (E) DMF/MeOH = 50/50 and (F) IPA/THF = 75/25. The ion peaks of Chimassorb 944 and Tinuvin 622 are depicted by inverted triangles and double circles, respectively. The number of polymer repeating units (x and y) and charges on the ions (z) are shown. DMF, N,N-dimethylformamide; IPA, isopropyl alcohol; LESA-MS, liquid extraction surface analysis-mass spectrometry; MeOH, methanol; THF, tetrahydrofuran; t-SPESI, tapping-mode scanning probe electrospray ionization. The spectrum data files are available in J-STAGE Data. https://doi.org/10.50893/data.massspectrometry.29146973 https://doi.org/10.50893/data.massspectrometry.29148194

Table 1. List of ion peaks derived from o-HALS.

m/z	x	y	z	Assigned ion	Name of o-HALS
993.951	1	—	1	[C24H49N4(C35H66N8)1H + H]+	Chimassorb 944
497.480	1	—	2	[C24H49N4(C35H66N8)1H + 2H]2+	Chimassorb 944
796.750	2	—	2	[C24H49N4(C35H66N8)2H + 2H]2+	Chimassorb 944
316.212	—	1	1	[H(C15H25NO4)1OCH3 + H]+	Tinuvin 622
599.390	—	2	1	[H(C15H25NO4)2OCH3 + H]+	Tinuvin 622
882.569	—	3	1	[H(C15H25NO4)3OCH3 + H]+	Tinuvin 622

o-HALS, oligomeric hindered amine light stabilizers; x and y the of repeating units of the monomers in Chimassorb 944 and Tinuvin 622; z, the charge number of the ion.

Next, the effects of the solvents on the ion signal intensities for t-SPESI-MSI were investigated. Notably, the mass spectra exhibit an opposite trend to that of the LESA results (Fig. 3E and 3F). When DMF/MeOH was used, the ions of Chimassorb 944 and Tinuvin 622 were detected; however, when IPA/THF was used, neither of the ions was detected.

A comparison of t-SPESI and LESA indicated that the physicochemical properties of the solvent required to improve ion detection sensitivity differed. In LESA, the extraction and ionization processes are separated; therefore, it is conceivable that the extraction and ionization efficiencies affect the ion signal intensity. The decreased intensity of the ion signal when utilizing DMF/MeOH can be explained by the challenges in consistently generating small charged droplets. This can be attributed to the high viscosity and high boiling temperature of DMF.

In t-SPESI, the solvent is supplied to the sample surface by an oscillating capillary probe, and extraction and ionization are performed at high speed. To utilize the dynamic flow of a small volume of solvent, it is necessary to hold the extract solution at the probe tip between the extraction and ionization events.

As supplemental information, in a previous study on the MSI of phospholipids in mouse brain tissue, the DMF/MeOH mixture was effective at increasing the signal intensity than DMF or MeOH alone.³¹⁾ DMF effectively increased the efficiency of lipid extraction, whereas MeOH effectively increased the amount of the extract solution at the probe tip and decreased the surface tension of the extract solution for ESI. Therefore, using a mixture of DMF and MeOH improved extraction and ionization. In addition, because the volume of the solvent used in t-SPESI is considerably small, it is conceivable that the ionization efficiency of ESI can be improved, similar to that of nano-ESI. Moreover, it is possible to concentrate the analytes extracted from the sample. Herein, the volume of solvent consumed in a single extraction–ionization event was ~0.25 pL when considering the solvent flow rate and the probe oscillation frequency.

The extraction–ionization process of t-SPESI requires the formation of liquid bridges, retention of the extractant at the probe tip, and ESI of the extractant all to be carried out sequentially. One conceivable explanation for the reduced signal intensity of o-HALS when using IPA/THF is that IPA and THF have lower surface tension, boiling point, and viscosity than DMF, which may have resulted in insufficient sampling and ionization.³²⁾ We are currently developing a method to directly observe the liquid bridge formed at the probe tip by incorporating an inverted microscope into a t-SPESI measurement system. This method is expected to facilitate the investigation of the relationship between the physicochemical properties of the solvent and the formation of liquid bridges.

Visualization of o-HALS and their photodegradation products in polyethylene film by t-SPESI-MSI

Figure 4A and 4B present the ion images for m/z 993.951 and 882.568, respectively. The structures of these ions were assigned as protonated molecules of Chimassorb 944 (x = 1) and Tinuvin 622 (y = 3) through subsequent LESA-MS/MS analysis (Fig. S1). In each ion image, the ion signal intensity was high in the nonirradiated region and decreased with increasing light irradiation time. Notably, the 2-ion images show an uneven distribution within the sample at the same ROI, including the unirradiated area, suggesting that o-HALS were unevenly distributed within the polyethylene film. This uneven distribution of these o-HALS may be attributed to the dry-blend film manufacturing process. Since the dry-blending method only allows the additive to adhere to the pellet surface, it is difficult to disperse it uniformly.³³⁾

Fig. 4. Ion images of protonated (A) Chimassorb 944 and (B) Tinuvin 622 obtained via t-SPESI. The white dashed lines in the ion images depict the boundary between the regions with different light irradiation times. The m/z values and the assigned chemical structures are shown below the ion images. (C) Relationship between light irradiation time and normalized ion signal intensity. t-SPESI, tapping-mode scanning probe electrospray ionization.

To compare the relative decrease in ion signal intensity, ROIs were defined as regions with different irradiation times, and the average and standard deviation of the ion signal intensities were calculated. The averaged signal intensities were normalized to those of the unirradiated areas (normalized ion signal intensity). Figure 4C presents the relationship between light irradiation time and normalized ion signal intensity. The intensity decreased nonlinearly with increasing irradiation time, suggesting that o-HALS degraded after light irradiation.

To investigate the degradation products of o-HALS formed early after light irradiation, PLS³⁴⁾ was used to select ions with a maximum signal intensity between 20 and 60 min of irradiation. Table S1 presents the top 10 selected ions with the highest PLS regression coefficients. Figure 5A–5C present the ion images of the 3 compounds (compound 1 [m/z 601.563], compound 2 [m/z 461.396], and compound 3 [m/z 629.558]). These were the top 3 compounds with the highest PLS regression coefficients. Moreover, these compounds were not detected under nonirradiated conditions but were detected after 5 min of irradiation, suggesting that they are degradation products of o-HALS generated by light irradiation. Figure 5D presents the relationship between light irradiation time and normalized ion signal intensity. Tables S2 and S3 show the correlation between the target variables set in PLS analysis and the ion signal intensities of compounds 1–3. Each ion showed different irradiation times for maximum signal intensity: compound 1 showed a maximum at 20 min, compound 2 showed a maximum at 60 min, and compound 3 showed a maximum signal intensity between 20 and 60 min. Notably, each ion showed a different profile, suggesting that the generation rate of the degraded product varied depending on the molecular structure. Furthermore, the degraded products remained in the film after light irradiation, suggesting that they were relatively stable under ambient conditions. Figure S2 presents the ion images of the remaining 7 compounds.

Fig. 5. Comparison of the changes in ion signal intensity as a function of light irradiation time. (A–C) Ion images of compounds whose signal intensities increased from 20 to 60 min. The m/z values and composition of each ion are shown. (D) Relationship between light irradiation time and normalized ion signal intensity. (E–G) Ion images of the compounds whose signal intensities increased after 180 min. (H) Relationship between the light irradiation time and averaged ion signal intensity. The white dashed lines in the ion images depict the boundary between regions with different light irradiation times.

Subsequently, the ions with the highest signal intensity after 360 min of light irradiation were selected. Table S2 presents the top 10 ions with the highest PLS regression coefficients. In addition, Fig. 5E–5H present the ion images of representative compounds (compound 4 [m/z 363.322], compound 5 [m/z 417.370], and compound 6 [m/z 403.354]) and the relationship between light irradiation time and ion signal intensity. Moreover, Fig. S4 presents the ion images of the other 7 compounds. All ions showed increased signal intensity after 180 min of irradiation. Tables S5 and S6 show the correlation between the target variables set in PLS analysis and the ion signal intensities of compounds 4–6.

Compounds 1–3 were degradation products generated by short-term light irradiation, and further photodegradation reactions likely produced compounds 4–6. Notably, compound 4 exhibited a relatively uniform distribution in the 180 min region, whereas compounds 5 and 6 showed a nonuniform distribution with high-intensity signal spots. This suggests that regions where photodegradation is promoted may be localized and potentially influenced by the local composition of the polymer film. The MSI results presented here indicate that o-HALS was unevenly distributed within the polyethylene film. However, it is also possible that the uneven distribution is attributed to changes in the ionization efficiency (ion suppression) caused by the coextracted components.

Structural estimation of o-HALS degradation products using LESA-MS/MS

To estimate the chemical structures of compounds 1–6, LESA-MS/MS was performed (Fig. 6). The product ion spectra showed the absence of fragment ions with mass differences of 22 Da (corresponding to the mass difference between Na⁺ adducts and proton adducts) and 17 Da (corresponding to the mass difference between ammonium adducts and proton adducts), indicating that all precursor ions were protonated molecules. The isotopic distribution of the ions suggests that they were monovalent. The molecular formulas of all the compounds were estimated based on their accurate masses. Furthermore, the experimental m/z values of all compounds showed errors <5 ppm compared with the theoretical values (Tables S1 and S2). Structural estimation was conducted using fragment ion information. Common fragment ions (m/z 58.065, 140.143, or 140.144) were detected in all mass spectra. These ions were identified as fragment ions (C₃H₈N⁺ and C₉H₁₈N⁺)³⁵⁾ derived from the characteristic tetramethylpiperidyl group of Chimassorb 944. Furthermore, fragment ions of 112.12 Da (C₈H₁₆), smaller than the precursor ions and corresponding to the formula weight of the tert-octyl group of Chimassorb 944, were detected. The estimated chemical structures of compounds 1–6 are as follows:

Fig. 6. LESA-MS/MS spectra of precursor ions of (A) m/z 601.563 (compound 1), (B) m/z 461.396 (compound 2), (C) m/z 629.558 (compound 3), (D) m/z 363.322 (compound 4), (E) m/z 417.370 (compound 5), and (F) m/z 403.354 (compound 6). The molecular structures corresponding to the product ions are indicated by arrows. LESA-MS/MS, liquid extraction surface analysis-tandem mass spectrometry. The spectrum data files are available in J-STAGE Data. https://doi.org/10.50893/data.massspectrometry.29146394

Compound 1: The precursor ion, m/z 601.563, was estimated to have the molecular formula [C₃₅H₆₈N₈ + H]⁺. This matched the proton adduct of the repeating unit of Chimassorb 944 (–C₃₅H₆₆N₈–), with hydrogen added to both ends. In addition, a fragment ion was detected at m/z 350.303 (C₂₇H₅₃N₈⁺), corresponding to the loss of the tert-octyl group (–C₈H₁₆) and the tetramethylpiperidyl group (–C₉H₁₈N) from the precursor ion. These results indicate that compound 1 had a structure corresponding to a monomer of Chimassorb 944 with the C–N bond between the triazine ring and the cleaved aminotetramethylpiperidine (Fig. 6A).

Compound 2: The precursor ion, m/z 461.396, was estimated to have the molecular formula [C₂₆H₄₈N₆O + H]⁺, indicating the presence of oxygen. The fragment ion m/z 210.135 (C₉H₁₆N₅O⁺) matched the formula of the precursor ion with the loss of the tert-octyl group (–C₈H₁₆) and the tetramethylpiperidyl group (–C₉H₁₈N), and the addition of hydrogen at the terminal. The absence of oxygenated ions related to the tetramethylpiperidyl group (m/z 58.065, 140.143, or 140.144) suggested that compound 2 did not have a nitroxide group with oxygen added to the amine of the tetramethylpiperidyl group but contained oxygen in other parts. Moreover, the molecular formula of compound 2 (C₂₆H₄₈N₆O) was derived from compound 1 (C₃₅H₆₈N₈) by the loss of C₉H₂₀N₂ and the addition of 1 oxygen, suggesting a structure in which the tetramethylpiperidyl group (–C₉H₁₈N₂) was cleaved from compound 1. The relative double bond value of compound 2 was 6, not 5, as expected, assuming it had a hydroxyl terminal. Furthermore, no neutral loss (18 Da) corresponding to the hydroxyl group was observed in the product ion spectrum, indicating the absence of a hydroxyl group. These considerations suggested that compound 2 has a structure wherein the C–N bond between the triazine ring and the aminotetramethylpiperidine is cleaved and the terminal is oxidized and cyclized with an ether bond (Fig. 6B).

Compound 3: The precursor ion, m/z 629.558, was estimated to have the molecular formula [C₃₆H₆₈N₈O + H]⁺, indicating the presence of oxygen. C₃₆H₆₈N₈O matched the molecular formula of compound 1 (C₃₅H₆₈N₈), with 1 terminal hydrogen replaced by an aldehyde group. For the same reasons as compound 2, compound 3 did not have a structure with a nitroxide group added to the amine of the tetramethylpiperidyl group but contained oxygen in other parts. Compound 3 has a structure in which the C–N bond between the triazine ring and aminotetramethylpiperidine was cleaved and oxidized to form an aldehyde group (Fig. 6C).

Compound 4: The precursor ion, m/z 363.322, was estimated to have the molecular formula [C₂₀H₃₈N₆ + H]⁺. The fragment ion m/z 112.062 (C₃H₆N₅⁺) matched the formula of the precursor ion with the loss of the tert-octyl (–C₈H₁₆) and the tetramethylpiperidyl (–C₉H₁₈N) groups and the addition of hydrogen at the terminal, corresponding to a fragment ion derived from diamino triazine, 251 Da smaller than the precursor ion. Compound 4 was estimated to have a structure in which the aminotetramethylpiperidyl and alkyl groups were cleaved from the structure of compound 1 (Fig. 6D).

Compound 5: The precursor ion at m/z 417.370 was estimated to have the molecular formula [C₂₄H₄₄N₆ + H]⁺. The fragment ion m/z 166.109 (C₇H₁₂N₅⁺) matched the formula of the precursor ion with the loss of the tert-octyl (–C₈H₁₆) and tetramethylpiperidyl (–C₉H₁₈N) groups and the addition of hydrogen at the terminal, corresponding to a fragment ion 251 Da smaller than the precursor ion, similar to compound 4. The difference in the molecular formulas between compounds 5 (C₂₄H₄₄N₆) and 4 (C₂₀H₃₈N₆) was C₄H₆, suggesting that it was an alkyl group attached to the aminotetramethylpiperidyl group and the structure was estimated, as shown in Fig. 6E.

Compound 6: The precursor ion, m/z 403.354, was estimated to have the molecular formula [C₂₃H₄₂N₆ + H]⁺. The fragment ion m/z 152.093 (C₆H₁₀N₅⁺) matched the formula of the precursor ion with the loss of the tert-octyl (–C₈H₁₆) and tetramethylpiperidyl (–C₉H₁₈N) groups and the addition of hydrogen at the terminal, corresponding to a fragment ion 251 Da smaller than the precursor ion, similar to compounds 4 and 5. The difference in the molecular formulas between compounds 6 and 4 (C₃H₄) suggested that it was an alkyl group, and compound 6 was estimated to have the structure shown in Fig. 6F.

Proposal of photodegradation reactions of Chimassorb 944

Multiple stepwise photodegradation reactions of Chimassorb 944 were proposed based on the results of t-SPESI-MSI and LESA-MS/MS (Fig. 7). The MSI results confirmed that the signal intensities of compounds 1–3 increased after a certain period of light irradiation. The MS/MS results for these ions indicate that the C–N bond of Chimassorb 944 was oxidatively cleaved to produce a monomer and that oxidation and cyclization reactions occurred at the end of the monomer. In addition, the signal intensities of compounds 4–6 increased owing to long-term light irradiation. It was presumed that the structures of these ions were produced by the cleavage of the alkyl chains of the monomer derivatives of compounds 1–3. Notably, in experiments in which Chimassorb 944 powder alone was irradiated with light, no photodegradation products were observed (data not shown), suggesting that the photodegradation of Chimassorb 944 was caused by radicals generated in the polymer.

Fig. 7. A possible multiple and stepwise photodegradation reaction of Chimassorb 944. Time indicates the light irradiation time at which the ion intensities of the compounds are at their maximum.

Notably, compounds 1–6 retained the tetramethylpiperidyl group, which is necessary for suppressing polymer photodegradation. However, it is unclear whether these photodegradation products effectively inhibit polymer degradation. In future, other analytical methods such as Fourier transform infrared spectroscopy, pyrolysis GC/MS, and crystallization temperature analysis with differential scanning calorimetry³⁶⁾ will be used to reveal structural changes in polymers, which will provide deeper insight into the relationship between o-HALS photodegradation and the polymer deterioration mechanism. Herein, the photodegradation products of Chimassorb 944 were analyzed, but those of Tinuvin 622 were not. We confirmed that the ion signal intensity of Tinuvin 622 (y = 3) decreased with increasing irradiation time (Fig. 4B), indicating photodegradation in the polymer film. Compared with Chimassorb 944, Tinuvin 622 does not contain primary or secondary amine groups. This may explain why the formation of protonated molecules in the photodegradation products was suppressed. Additionally, alternative ionization strategies such as the addition of sodium or ammonium sources to the solvent to promote the formation of adduct ions may be useful for detecting Tinuvin 622 photodegradation products.

The aforementioned results shed light on the variations in the ion image and ion structures with respect to the light irradiation time. Based on our current understanding, this study appears to be a pioneering effort to reveal that the photodegradation of Chimassorb 944 within a polyethylene film involves a complex series of chemical reactions.

CONCLUSION

We used t-SPESI-MSI and LESA-MS/MS to investigate the photodegradation products of o-HALS in the polyethylene films. The distribution of Chimassorb 944 and Tinuvin 622 clearly diminished after light irradiation. The photodegradation products of Chimassorb 944 were found to depend on the irradiation time. The structural estimation of the products suggested the fragmentation of Chimassorb 944 at specific chemical bonds, whereas the tetramethylpiperidyl group, which is necessary for suppressing polymer photodegradation, was retained. The combination of direct extraction and ionization methods provided us with comprehensive information about multiple and stepwise photodegradation processes. We believe that our proposed approaches will provide insights into additive photodegradation and potentially contribute to the development of light-resistant polymer materials.

DATA AVAILABILITY STATEMENT

The spectrum data files of Figs. 3, and 6 are available in J-STAGE Data.

ACKNOWLEDGMENTS

This work was partially supported by JSPS KAKENHI grant number JP23H03711.

Notes

Mass Spectrometry (Tokyo) 2025; 14(1): A0173

REFERENCES

• 1)  E. Yousif, R. Haddad. Photodegradation and photostabilization of polymers, especially polystyrene, Review. SpringerPlus 2: 398, 2013.
• 2)  Y. Ohkatsu. Search for unified action mechanism of hindered amine light stabilizers. J. Jpn. Petrol. Inst. 51: 191–204, 2008.
• 3)  M. R. L. Paine, P. J. Barker, S. J. Blanksby. Desorption electrospray ionisation mass spectrometry reveals in situ modification of a hindered amine light stabiliser resulting from direct N–OR bond cleavage. Analyst 136: 904–912, 2011.
• 4)  J. L. Hodgson, M. L. Coote. Clarifying the mechanism of the Denisov cycle: How do hindered amine light stabilizers protect polymer coatings from photo-oxidative degradation? Macromolecules 43: 4573–4583, 2010.
• 5)  E. T. Denisov. Mechanism of regeneration of hindered nitroxyl and aromatic amines. Polym. Degrad. Stabil. 25: 209–215, 1989.
• 6)  E. T. Denisov. The role and reactions of nitroxyl radicals in hindered piperidine light stabilisation. Polym. Degrad. Stabil. 34: 325–332, 1991.
• 7)  P. Gijsman, A. Dozeman. Comparison of the UV-degradation chemistry of unstabilized and HALS-stabilized polyethylene and polypropylene. Polym. Degrad. Stabil. 53: 45–50, 1996.
• 8)  B. Forsthuber, G. Grüll. The effects of HALS in the prevention of photo-degradation of acrylic clear topcoats and wooden surfaces. Polym. Degrad. Stabil. 95: 746–755, 2010.
• 9)  C. W. Klampfl, M. Himmelsbach. Advances in the determination of hindered amine light stabilizers - A review. Anal. Chim. Acta 933: 10–22, 2016.
• 10)  J. Malík, A. Hrivík, E. Tomová. Diffusion of hindered amine light stabilizers in low density polyethylene and isotactic polypropylene. Polym. Degrad. Stabil. 35: 61–66, 1992.
• 11)  J. Malík, A. Hrivík, D. Alexyová. Physical loss of hindered amine light stabilizers from polyethylene. Polym. Degrad. Stabil. 35: 125–130, 1992.
• 12)  W. Buchberger, M. Stiftinge. Analysis of polymer additives and impurities by liquid chromatography/mass spectrometry and capillary electrophoresis/mass spectrometry. in Mass Spectrometry of Polymers – New Techniques (Ed: M. Hakkarainen), Springer Berlin Heidelberg, Berlin, Heidelberg, 2012, pp. 39–67.
• 13)  R. Trones, T. Andersen, T. Greibrokk, D. R. Hegna. Hindered amine stabilizers investigated by the use of packed capillary temperature-programmed liquid chromatography. I. Poly((6-((1,1,3,3-tetramethylbutyl)-amino)-1,3,5-triazine-2,4-diyl)-(2,2,6,6-tetramethyl-4-piperidyl)imino)-1,6-hexanediyl ((2,2,6,6-tetramethyl-4-piperidyl)imino)). J. Chromatogr. A 874: 65–71, 2000.
• 14)  R. Noguerol-Cal, J. M. López-Vilariño, M. V. González-Rodríguez, L. Barral-Losada. Effect of several variables in the polymer toys additive migration to saliva. Talanta 85: 2080–2088, 2011.
• 15)  M. Gill, M. J. Garber, Y. Hua, D. Jenke. Development and validation of an HPLC-MS-MS method for quantitating bis(2,2,6,6-tetramethyl-4-piperidyl) sebacate (Tinuvin 770) and a related substance in aqueous extracts of plastic materials. J. Chromatogr. Sci. 48: 200–207, 2010.
• 16)  M. Reisinger, S. Beissmann, W. Buchberger. Quantitation of hindered amine light stabilizers in plastic materials by high performance liquid chromatography and mass spectrometric detection using electrospray ionization and atmospheric pressure photoionization. Anal. Chim. Acta 803: 181–187, 2013.
• 17)  C. Schwarzinger, S. Gabriel, S. Beissmann, W. Buchberger. Quantitative analysis of polymer additives with MALDI-TOF MS using an internal standard approach. J. Am. Soc. Mass Spectrom. 23: 1120–1125, 2012.
• 18)  S. T. Hsiao, M. C. Tseng, Y. R. Chen, G. R. Her. Analysis of polymer additives by matrix-assisted laser desorption ionization/time of flight mass spectrometer using delayed extraction and collision induced dissociation. J. Chin. Chem. Soc. (Taipei) 48(6A): 1017–1027, 2001.
• 19)  K. D. Jansson, C. P. Zawodny, T. P. Wampler. Determination of polymer additives using analytical pyrolysis. J. Anal. Appl. Pyrolysis 79: 353–361, 2007.
• 20)  Y. Taguchi, Y. Ishida, S. Tsuge, H. Ohtani, K. Kimura, T. Yoshikawa, H. Matsubara. Structural change of a polymeric hindered amine light stabilizer in polypropylene during UV-irradiation studied by reactive thermal desorption-gas chromatography. Polym. Degrad. Stabil. 83: 221–227, 2004.
• 21)  M. Haunschmidt, C. W. Klampfl, W. Buchberger, R. Hertsens. Rapid identification of stabilisers in polypropylene using time-of-flight mass spectrometry and DART as ion source. Analyst 135: 80–85, 2010.
• 22)  S. M. Reiter, W. Buchberger, C. W. Klampfl. Rapid identification and semi-quantitative determination of polymer additives by desorption electrospray ionization/time-of-flight mass spectrometry. Anal. Bioanal. Chem. 400: 2317–2322, 2011.
• 23)  M. R. L. Paine, G. Gryn’ova, M. L. Coote, P. J. Barker, S. J. Blanksby. Desorption electrospray ionisation mass spectrometry of stabilised polyesters reveals activation of hindered amine light stabilisers. Polym. Degrad. Stabil. 99: 223–232, 2014.
• 24)  L. Yang, L. He, J. Xue, Y. Ma, Z. Xie, L. Wu, M. Huang, Z. Zhang. Persulfate-based degradation of perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) in aqueous solution: Review on influences, mechanisms and prospective. J. Hazard. Mater. 393: 122405, 2020.
• 25)  M. R. Paine, P. J. Barker, S. A. Maclauglin, T. W. Mitchell, S. J. Blanksby. Direct detection of additives and degradation products from polymers by liquid extraction surface analysis employing chip-based nanospray mass spectrometry. Rapid Commun. Mass Spectrom. 26: 412–418, 2012.
• 26)  R. Shimazu, Y. Yamoto, T. Kosaka, H. Kawasaki, R. Arakawa. Application of tapping-mode scanning probe electrospray ionization to mass spectrometry imaging of additives in polymer films. Mass Spectrom. (Tokyo) 3: S0050, 2014.
• 27)  I. Hintersteiner, L. Sternbauer, S. Beissmann, W. W. Buchberger, G. M. Wallner. Determination of stabilisers in polymeric materials used as encapsulants in photovoltaic modules. Polym. Test. 33: 172–178, 2014.
• 28)  Y. Otsuka, M. Okada, T. Hashidate-Yoshida, K. Nagata, M. Yamada, M. Goto, M. Sun, H. Shindou, M. Toyoda. Improved ion detection sensitivity in mass spectrometry imaging using tapping-mode scanning probe electrospray ionization to visualize localized lipids in mouse testes. Anal. Bioanal. Chem. 417: 275–286, 2025.
• 29)  Y. Otsuka, B. Kamihoriuchi, A. Takeuchi, F. Iwata, S. Tortorella, T. Matsumoto. High-spatial-resolution multimodal imaging by tapping-mode scanning probe electrospray ionization with feedback control. Anal. Chem. 93: 2263–2272, 2021.
• 30)  L. L. Lopez, P. R. Tiller, M. W. Senko, J. C. Schwartz. Automated strategies for obtaining standardized collisionally induced dissociation spectra on a benchtop ion trap mass spectrometer. Rapid Commun. Mass Spectrom. 13: 663–668, 1999.
• 31)  Y. Otsuka, N. Ote, M. Sun, S. Shimma, O. Urakawa, S. Yamaguchi, T. Kudo, M. Toyoda. Solvent effects of N,N-dimethylformamide and methanol on mass spectrometry imaging by tapping-mode scanning probe electrospray ionization. Analyst 148: 1275–1284, 2023.
• 32)  D. C. Han, Y. J. Jin, J. H. Lee, S. I. Kim, H. J. Kim, K. H. Song, G. Kwak. Environment-specific fluorescence response of microporous, conformation-variable conjugated polymer film to water in organic solvents: On-line real-time monitoring in fluidic channels. Macromol. Chem. Phys. 215: 1068–1076, 2014.
• 33)  M. Alonso, F. J. Alguacil. Dry mixing and coating of powders. Rev. Metal. 35: 315–328, 1999.
• 34)  Y. Ilin, M. L. Kraft. Secondary ion mass spectrometry and Raman spectroscopy for tissue engineering applications. Curr. Opin. Biotechnol. 31: 108–116, 2015.
• 35)  M. R. L. Paine, P. J. Barker, S. J. Blanksby. Characterising in situ activation and degradation of hindered amine light stabilisers using liquid extraction surface analysis-mass spectrometry. Anal. Chim. Acta 808: 190–198, 2014.
• 36)  N. M. Ainali, D. N. Bikiaris, D. A. Lambropoulou. Aging effects on low- and high-density polyethylene, polypropylene and polystyrene under UV irradiation: An insight into decomposition mechanism by Py-GC/MS for microplastic analysis. J. Anal. Appl. Pyrolysis. 158: 105207, 2021.