Enhancing mass analysis of ultra-high molecular weight polystyrene: a comparative study of copper and silver salts with MALDI mass spectrometry

Abstract

This study presents a report on the use of copper (Cu) salts in polymer mass spectrometry for samples exceeding several thousand daltons, demonstrating their efficiency and cost-effectiveness for analyzing ultra-high molecular weight (UHMW) polystyrene (PS). Using matrix-assisted laser desorption/ionization (MALDI) coupled with a home-built linear ion trap mass spectrometer (LIT-MS) under optimized conditions, polystyrene (PS) with masses up to two million daltons was successfully detected. This method using MALDI LIT-MS surpasses the previous upper detection limits observed with silver (Ag) and cesium (Cs) salts using a DCTB matrix. In the MALDI time-of-flight (TOF) mass spectrometry system, these salts were ineffective for detecting PS masses in the range from 660 kDa to 1.1 million Da, respectively (Rapid Commun. Mass Spectrom., 2015, 29, 1039–1046). In this study, we demonstrate for the first time that both Ag and Cu salts can successfully analyze UHMW PS samples up to 2 million Da using the trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene]-malononitrile (DCTB) matrix. However, Ag salts predominantly generate multiply charged state ion signals with no control over charge states. In contrast, Cu salts demonstrate superior versatility in modulating charge states, with copper(I) salts achieving higher charge states of up to +5 and copper(II) salts generating lower charge states of up to +3. This targeted control simplifies analysis and enhances the method's flexibility, making Cu salts particularly valuable for analyzing both individual PS samples and mixtures. Additionally, Cu salts are compatible with both DCTB and all-trans-retinoic acid (RA) matrices, with DCTB delivering superior performance. Comparative analysis with Ag salts further validates Cu salts as optimal cationizing reagents for UHMW PS analysis, offering a powerful tool for mass distribution measurements in polymer research and advancing analytical methods in this field.

---

Introduction

In recent years, advances in polymer synthesis have led to the development of highly complex polymeric materials, including intricate oligomer mixtures, polymers with diverse shapes, and those with ultra-high molecular weights (UHMW).^1–3 Understanding the mass distribution, structure, and composition of UHMW polymers is crucial for optimizing their performance in terms of tensile strength, thermal properties, chemical resistance, and processing techniques.^1,4 In addition to synthesizing UHMW polymers, significant advancements have been made in the mass distribution analysis of these UHMW samples, employing chromatography techniques,^5,6 light scattering techniques,⁷ and the matrix-assisted laser desorption/ionization (MALDI) mass spectrometry technique.^8–10 Size-exclusion chromatography (SEC) is a frequently employed method for polymer separation and mass analysis.⁵ MALDI coupled with a time-of-flight (TOF) analyzer is a suitable method for UHMW polymer mass analysis, providing molecular weight distribution⁸ and information about oligomer aggregates.⁹ However, analyzing UHMW polymers using MALDI poses challenges related to ion detection and ionization efficiency. Recent advancements in high-mass detectors, such as the superconducting tunnel junction (STJ) cryodetector⁹ and commercial high acceleration voltage secondary electron detector,¹¹ have addressed some detection limitations of UHMW polymers. Additionally, charge detector coupled to linear ion trap mass analyzer has shown the potential for the ionization of UHMW polymer samples using MALDI.¹² Despite these developments in high-mass ion detection, challenges in the ionization of high-mass polymer samples still limit the use of MALDI-MS for UHMW polymer mass analysis. The existing sample preparation protocols, encompassing matrix selection, cationization reagent choice, soluble solvent considerations, etc., have not been adequately optimized for high-mass polymer analysis as compared to their low-mass counterparts. This lack of optimization contributes to poor ionization efficiency, which adequately limits the use of MALDI-MS in routine UHMW PS analysis.

The choice of a suitable matrix and cationization reagent is critical in determining the success of MALDI analysis of polymers. For polymers, it is critical to use matrices that are compatible with the sample and provide optimal ionization efficiency.^13–15 The type of polymer affects the choice of the matrix and cationization reagent for mass analysis. Different matrices are suitable for polar and nonpolar polymers.^15,16 In the context of nonpolar polymers such as polystyrene, various MALDI matrices such as dithranol, 2,5-dihydroxybenzoic acid (DHB), iodoacetic acid (IAA), all-trans-retinoic acid (RA), 2,4,6-trihydroxyacetophenone (THAP), trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene]malononitrile (DCTB) and others have been used.^{9,11,13,15,17,18} Similarly, the selection of an appropriate cationization reagent for the MALDI sample preparation of the analyte under investigation is important. Transition metal cations such as Cu²⁺, Zn²⁺, Co²⁺, Au³⁺, Al⁺, Fe³⁺, Ni²⁺, and Pd^{2+15,16,19–23} as well as alkali metal cations such as Cs⁺, Na⁺, K⁺, Rb⁺ and Li^{+19,21,24–26} have been used to facilitate the generation of low-mass polymer ions with MALDI.^{14,19,20,27,28}

The use of Cu salts as cationization reagents in polymer mass analysis has been explored, though less extensively compared to Ag salts. Cu salts can be utilized with both polar and nonpolar matrices for analyzing low molecular weight polymers up to several thousand kDa.^15,16 Macha et al.¹⁵ demonstrated that nonpolar matrices such as anthracene, pyrene, and acenaphthene are as effective as conventional acidic matrices in MALDI-MS analysis of synthetic polymers such as polybutadiene, polyisoprene, and polystyrene. Notably, the effectiveness of Cu salts varied with the polymer type, whereas Ag salts were consistently effective across all polymers. In contrast, Habumugisha et al.¹⁶ highlighted a significant issue with Ag salts, which tend to form interfering clusters when used with polar matrices in MALDI TOF-MS analyses of polystyrene nanoplastics. This interference, however, was effectively reduced by employing Cu(II) chloride (CuCl₂), illustrating an advantage of Cu salts in reducing signal noise. Furthermore, Yalcin et al.²³ further demonstrated the benefits of using Cu(II) nitrate (Cu(NO₃)₂) with the RA matrix in MALDI methods, significantly enhancing ionization performance and extending the upper mass detection limits of polymers. This approach enabled the analysis of polybutadienes and polyisoprenes with narrow polydispersity, achieving mass detection up to 300 000 Da and 150 000 Da, respectively.

Amongst the various matrices and salts discussed above, only two matrices, RA and DCTB, in combination with Ag⁺, Rb⁺, and Cs⁺ salts, have been successfully employed for the generation of UHMW PS ions.^8,9,11 Schriemer and Li reported that multiply charged ions (from +2 to +4) of PS with a mass of ∼1.5 MDa were generated using Ag⁺ cations and the RA matrix.⁸ Aksenov and Bier measured the PS aggregates (∼4 MDa) with charges ≥+2 using the RA matrix and Ag salts.⁹ Moreover, Gabriel et al. investigated the ionization of PS ions with Ag⁺, Rb⁺, and Cs⁺ cations, and with the DCTB matrix.¹¹ Their study concluded that Ag salts are not suitable for UHMW mass analysis with a DCTB matrix, which is surprising given their successful use in cationizing UHMW polymers with an RA matrix in other studies. Although the CsCl salt was proposed as a more effective alternative, only poor PS ion signals (S/N ≤ 3) were observed using a DCTB matrix, specifically for masses of 650 kDa with Rb⁺ and Cs⁺ cations and 1.1 MDa with Cs⁺ cations only. Moreover, no PS ions could be generated using Ag⁺ cations for PS samples heavier than 240 kDa. Given the inability of Ag salts to analyze PS above 240 kDa and the limited ionization efficiency of UHMW PS with Cs salts, it is crucial to explore additional salts for use with the DCTB matrix.

In this study, we investigated alternative cationization reagents, i.e., Cu salts, to enhance ionization efficiency for UHMW PS polymers with a home-built MALDI LIT-MS.¹² Moreover, to align with prior findings that Ag salts could successfully cationize UHMW PS using an RA matrix but failed with a DCTB matrix, we included Ag salts with DCTB in the present study and directly compared them with Cu salts under identical experimental conditions. This approach provides a comprehensive evaluation of the effectiveness of both salts in UHMW PS analysis and ensures our results remain consistent with existing UHMW polymer mass spectrometry research. Our results demonstrate that Cu⁺ cations produce PS ions with lower and higher charge-state distributions depending on the choice of salts and significantly improve ion S/N ratios ranging from ∼30 to 100. Cu salts are suitable cationization salts for both RA and DCTB matrices in generating UHMW PS ions. The DCTB matrix, however, shows a higher ionization efficiency for these polymers as compared to that with the RA matrix. Our study demonstrates that Cu salts offer a distinct advantage in modulating the charge states of PS ions via the changes in metal oxidation states of Cu salts, allowing for greater flexibility in charge state control as compared to the use of Ag salts. Specifically, copper(II) salts predominantly produced PS ions with charge states up to +3, while copper(I) salts enabled higher charge states, reaching up to +5. In contrast, silver salts provide limited salt options and mostly generate multiply charged PS ions up to five charge states. The finding of Ag salts generating UHMW PS ions using the DCTB matrix contradicts a previous study that claimed incompatibility between Ag salts and the DCTB matrix in the analysis of UHMW PS samples.¹¹ Furthermore, the alkali metal caesium (Cs) salt was found ineffective for producing PS ions >606 kDa, underscoring its limitations compared to Ag and Cu salts. Overall, Cu salts are identified as cost-effective substitutes for Ag salts in the analysis of UHMW PS samples up to 2 million daltons using MALDI mass spectrometry.

Experimental section

Experimental setup

Mass measurements were performed with a homebuilt linear ion trap mass spectrometer (LIT-MS) equipped with a charge sensing particle detector (CSPD) for high mass ion detection. The instrument design has already been described elsewhere.¹² In brief, the ionization and mass analyzer chambers were separated, and ions were transmitted using a quadrupole ion guide (Q1) placed between the ionization and mass analyzer chambers. A commercial three-segment LIT (Thermo-Fisher Scientific, Waltham, MA, USA) was used as the mass analyzer. The charge-sensing particle detector (CSPD)²⁹ was arranged radially to detect ions ejected radially using frequency scanning with dipolar resonance ejection. The details of frequency scanning with dipolar resonance ejection have already been described elsewhere.¹² The MALDI sample was ionized using a 349 nm Nd:YLF laser (Model: Explorer One 349, Spectra-Physics, MKS Instruments, Inc., MA, USA). The pulse duration was 35 ns, with a pulse repetition rate of 100 Hz and a laser repetition rate of 300 shots. The UV laser was aimed at the target probe and the laser beam passed through the LIT and the Q1. The sample was prepared on a stainless-steel sampling probe and then transferred using a rod to the Q1 chamber. The laser hit along the surface normal to the sample surface. Ions were guided to the LIT using a quadrupole ion guide and trapped axially using a DC potential applied to the end cap lens of the LIT mass analyzer. For collisional cooling and ion trapping of UHMW polystyrene ions, buffer gas molecules of helium (He) and argon (Ar) were used separately in the ionization and LIT regions. To optimize mass analysis, He gas was introduced near the ionization region for collisional cooling, and Ar gas was used in the LIT chamber for efficient ion trapping. The He gas pressure was set to ∼28 mTorr, and the Ar gas pressure in the LIT chamber was kept constant at 0.6 mTorr. Finally, the trapped ions were ejected radially by sweeping both RF and AC frequencies while the amplitude of RF and AC waveforms were kept constant and resonance ejection was achieved at fixed β_eject = 0.8. The experimental conditions for acquiring different PS mass spectra with the LIT mass spectrometer are listed in Table S1,† including the RF frequencies and voltages applied to the Q1 and LIT, as well as the AC frequencies and voltages applied to the X pair of LIT electrodes. MATLAB software was used as an interface to collect the data which were processed by wavelet analysis and a numerical rectifier to remove the strong noise interference from the radio frequency and amplitude applied to the LIT mass analyzer. The details of noise cancellation are provided elsewhere.¹² Origin 9 software was used to plot mass spectra and perform baseline correction. Five data sets were collected to plot each mass spectrum and an error bar was depicted in each figure. The laser pulse energy was kept constant at 30 μJ throughout the experiments, and all mass spectra were obtained with this laser fluence using the DCTB matrix. In comparison to the DCTB matrix, the RA matrix required ∼15% more laser energy for PS mass analysis. All other experimental parameters such as guiding quadrupole settings, chamber pressures, etc. were optimized according to each PS sample and kept constant for each PS mass analysis. The only variable was the use of different metal cationization reagents.

Samples and sample preparation

In an effort to optimize matrix selection for UHMW PS mass analysis, we explored matrices such as RA and DCTB, which were previously used using Ag salts^8,9 and Cs salts,¹¹ respectively. In addition, we assess the effectiveness of several Cu salts for cationizing UHMW PS samples. Three cations, i.e., Ag⁺, Cu⁺, and Cs⁺, were adopted to examine the cationization of UHMW PS samples. The silver salts, e.g., silver nitrate (AgNO₃) and silver trifluoroacetate (AgTFA), the copper salts, e.g., copper(I) chloride (CuCl), copper(II) chloride (CuCl₂), copper(I) bromide (CuBr), copper(II) bromide (CuBr₂), copper(I) iodide(CuI), and copper(II) acetate (Cu(OAc)₂), and the alkali metal cesium chloride (CsCl) salt were used in this study. Polystyrene of PS210k (M_w ∼ 240 kDa), AgNO₃ (>99.0%), and AgTFA (>98.0%) were purchased from Sigma-Aldrich (St. Louis, Missouri, USA). Polystyrenes of PS650k (M_w ∼ 606 kDa), PS900k (M_w ∼ 946 kDa), and PS2M (M_w ∼ 2000 kDa) were purchased from Pressure Chemicals Co. (St, Pittsburgh, PA, USA). Trans-2-[3-(4-tert-butylphenyl)-2-methylprop-2-enylidene]-malononitrile (DCTB), CsCl (>99.0%), CuI (>98.0%), CuBr (>98.0%), CuBr₂ (>98.0%), and CuCl₂ (>98.0%) were purchased from Tokyo Chemical Industry (Tokyo, Japan). Cu(OAc)₂ (>98.0%), tetrahydrofuran (THF), and acetonitrile (ACN) were purchased from J. T. Baker (Phillipsburg, NJ, USA). The THF solvent was used to prepare all PS, DCTB matrix, and cationization reagent stock solutions except for CuCl, CuCl₂, CuBr, and CuBr₂. The CuCl, CuCl₂, CuBr, and CuBr₂ were prepared in ACN solution. Using the salts immediately after sample preparation is essential because any delay could decrease ion signal intensities or cause experimental failure. Moreover, it is necessary to store unused Cu(I) salts in inert gas because the stability of Cu(I) salts is time-dependent and Cu(I) salts are prone to disproportionation reactions in the presence of oxygen and moisture, resulting in the formation of Cu(0) and Cu(II).^30,31

The matrix solution was prepared at a constant concentration of 25 mg mL⁻¹ (0.1 M). The polystyrene standards used in this study were prepared with concentrations ranging from 5 × 10⁻⁵ M to 5 × 10⁻⁷ M. Cationization reagents were prepared in THF and ACN at a concentration of 0.15 M and sonicated for 20 minutes before use. The PS, matrix, and cationization reagent were mixed in a microcentrifuge tube with a volume ratio of 3 : 110 : 2 (v/v/v) for PS210k and 3 : 200 : 1 (v/v/v) for high-mass polystyrenes unless otherwise specified. A 3–4 μL mixture was placed on a stainless-steel sampling probe (8 mm in diameter), and homogeneous samples were prepared using the forced-dried droplet (FDD) method.^32,33 This approach involves careful agitation of the droplet on the probe surface with a pipette tip, ensuring an even and thin layer across the target. The FDD method has been previously validated for producing uniform protein and polymer microcrystals;^12,32–34 therefore, the FDD method was adopted to prepare a homogeneous PS sample surface in this study.

Ni et al.³⁵ suggested that the concentration of cations in the gas phase was significantly influenced by the even distribution of cationization salt crystals within the matrix and the strength of the interaction between these salt crystals and the matrix crystals. By employing the FDD method in the present study, it was assumed that the distribution of cation salts within the MALDI sample crystals is uniform to facilitate cation adduct formation.^32,33 The thin-layer sample surface prepared by FDD could result in efficient ionization of the UHMW PS samples, resulting in enhanced ion signal intensities and reduced ion signal variations.

Calibration

Calibration of the MALDI LIT-MS from 100 kTh to 2 MTh was challenging because calibrants for such a high mass and for such a broad mass range were not available. To address this, we utilized the tetrameric form of the alpha-2-macroglobulin (alpha2M) protein standard, with a molecular weight of 725 kDa, for calibration of mass spectra of PS210k and PS650k (Fig. S1a†). The dimeric form of alpha2M provided multiply charged ion signals, allowing us to use the singly (∼362.5 kTh), doubly (∼181.2 kTh), and triply (∼121 kTh) charged ions for a three-point calibration. Mass spectra for PS210k and PS650k with AgNO₃ were subsequently obtained under similar experimental conditions of alpha2M. Following calibration, the mass peaks for PS210k and PS650k were observed at approximately 240 kTh and 606 kTh, respectively. PS900k was calibrated further by acquiring mass spectra with a mixture of PS650k and PS900k samples (Fig. S1b†) and the M_w of PS900k was assigned to ∼946 kDa. The values of M_w and M_n were obtained using gel permeation chromatography (GPC) and light scattering by the suppliers, and are summarized in Table S2.†

Results and discussion

Optimal polystyrene-to-matrix ratio in the absence of cationization reagents

The PS ion signals could be generated with the DCTB matrix only in the absence of cationization reagents because sodium (Na) and potassium (K) are the impurities of the DCTB matrix.^11,22 The optimal PS-to-matrix (PS/M) ratio was examined to acquire PS ion signals. In Fig. S2,† PS210k was analyzed at various PS : DCTB matrix concentrations (v/v), i.e., 3 : 50, 3 : 100, 3 : 150, 3 : 200, and 3 : 250. Notably, strong singly charged monomer (∼240 kDa) and weak dimer ion signals were observed. The ion signal intensity of the monomer was strongly dependent on the PS/M concentration ratio. The ion signal intensity increased as the PS/M ratio increased from 3 : 50 to 3 : 150 (Fig. S2a–c†) and decreased as the PS/M ratio increased to 3 : 200 and 3 : 250 (Fig. S2d and e†). The optimal PS/M ratio was observed at 3 : 150. The reduction in signal intensity was a result of the excess matrix suppressing the analyte ionization and reducing vacant coordination sites for analyte incorporation into crystals. In addition, the formation of matrix clusters and the uneven distribution of clusters across the sample might restrict uniform laser energy absorption. Although the cation adducts (Na, K, etc.) available in the DCTB matrix can help get PS ion signals, they are insufficient to ionize PS with a mass over 606 kDa (data not shown). This can be solved by adding cationization reagents, which boost the attachment of cations to PS in the gas phase.^36–40

The optimal ratio of polystyrene : matrix : AgNO₃ cationization reagent

In general, the addition of cationization reagents (CAT) is adopted to increase signal intensities of PS ions. Here we selected the AgNO₃ salt and added it into the PS and DCTB matrix solution to find the optimal PS/M/CAT ratio according to the method proposed by Wang et al.⁴¹ To reduce the complexity, the DCTB matrix and AgNO₃ cationization reagent concentrations remained constant at 0.1 M and 0.15 M and only PS concentrations were varied. As shown in Fig. 1, different concentrations of three PS samples with molecular weights of 240 kDa, 606 kDa, and 946 kDa were investigated to obtain the maximum PS ion signals. As shown in Fig. 1a, the concentration range of the PS210k sample was from 10 μM to 500 μM. The signal intensities of +1 ions increased as the PS concentrations decreased from 500 μM to 50 μM and decreased as the concentration decreased from 50 μM to 10 μM. From the combined intensities of the +1 and +2 charge states, it appears that a broader concentration range (10–150 μM) is optimal for the mass analysis of PS210k. As shown in Fig. 1b, the concentration range of the PS650k sample was from 1 μM to 100 μM. The signal intensities of +1, +2, and +3 ions increased as the PS concentration decreased from 100 μM to 10 μM and decreased as the concentration reduced from 10 μM to 1 μM, indicating that the optimal PS concentration for PS650k is 10 μM. As shown in Fig. 1c, the concentration range of the PS900k sample was from 0.1 μM to 10 μM. The signal intensities of +1, +2, +3, and +4 ions increased as the PS concentration decreased from 10 μM to 5 μM and decreased as the concentration reduced from 5 μM to 0.1 μM, suggesting that the optimal PS concentration for PS900k is 5 μM. It is noted that as the PS mass increases the optimal concentration range becomes narrower, reflecting the distinct ionization behavior of heavier molecules. The optimal PS : DCTB : AgNO₃ molar ratios for PS210k, PS650k, and PS900k samples were calculated to be 3 : 110 : 2, 3 : 200 : 1, and 3 : 200 : 1 (Table 1), and Fig. S3† shows the corresponding mass spectra. Fig. 1 shows that the PS : matrix : CAT ratio has a significant influence on signal intensity across all charge states while having no effect on charge state distributions. The charge state distributions might be related to the properties of the salts and not the PS : matrix : CAT ratios. Therefore, mass analyses of various UHMW PS samples were performed to determine the effect of salts on the charge state distribution of PS samples using different salts with varying properties such as lattice energy, cation metal oxidation states, etc.

Fig. 1 The effect of PS concentration on polymer ionization efficiency of (a) PS210k, (b) PS650k, and (c) PS900k. To optimize the PS : matrix : cationizing reagent ratio, the PS concentration is varied and mixed with 0.1 M DCTB and 0.15 M AgNO₃.

Table 1 Optimized MALDI sample preparation conditions for UHMW PS samples using the DCTB matrix (0.1 M) and AgNO₃ cationization reagent (0.15 M)

Polystyrene		Volume (μL) in mixture			Molar ratio
	Concentration (molar)	PS	DCTB	AgNO3	DCTB : PS	AgNO3 : PS
PS210k	5.0 × 10^−5	3	110	2	73 231	1962
PS650k	1.0 × 10^−5	3	200	1	665 735	4905
PS900k	5.0 × 10^−6	3	200	1	1 331 469	9811

---

The effect of the metal in +1 and +2 oxidation states on polystyrene mass spectra

It was observed that the metal oxidation state of transition metal salts might affect the charge state distribution and signal intensities of PS ions. Fig. 2 shows mass spectra of PS650k with the DCTB matrix using AgTFA, AgNO₃, CsCl, CuI, CuBr, and CuCl salts with the metal in the +1 oxidation state. With AgTFA salt, up to +5 charge states were observed, and the signal intensity of +2 ions was the highest (Fig. 2a). With AgNO₃ salt, the charge state of ions reduced to +3, and the signal intensity of +1 ions was the highest (Fig. 2b). With the alkali metal CsCl salt, the charge state reduced to +2, and the signal intensity of +1 ions was the highest while the signal intensity of +2 ions decreased significantly (Fig. 2c). With CuI salt, the charge state of PS650k ions increased up to +4, and the signal intensities of +2 and +1 ions were the highest (Fig. 2d). Similarly, with CuBr salt, a charge state up to +4 was observed (Fig. 2e), but the signal intensity of +4 ions was lower than that observed with CuI (Fig. 2d). With CuCl salt, the charge state reduced to +3, and the charge state distribution was comparable to that obtained with AgNO₃ salt and the DCTB matrix.

Fig. 2 Mass spectra of PS650k obtained with a DCTB matrix using various cationizing reagents with the metal in the +1 oxidation state. The cationizing reagents used were (a) AgTFA, (b) AgNO₃, (c) CsCl, (d) CuI, (e) CuBr, and (f) CuCl. The optimal ratio of PS : matrix : cationizing reagent was 3 : 200 : 1 (v/v/v).

Mass spectra of PS650k ions were acquired with CuBr₂, CuCl₂, and Cu(OAc)₂ salts, and their Cu metal was in the +2 oxidation state, as shown in Fig. 3. As shown in Fig. 3a, the signal intensities of +2 and +3 ions with Cu(II)Br₂ salt became weaker than those with Cu(I)Br salt. As shown in Fig. 3b, the ion signal intensities reduced with Cu(II)Cl₂ salt as compared to that with Cu(I)Cl salt. In Fig. 3c, the signal intensities of +2 and +3 ions with Cu(OAc)₂ salt were about a factor of two higher than those with CuBr₂ salt and CuCl₂ salt. The above findings suggested that the metal oxidation states of Cu salts (Cu(I) and Cu(II)) do influence the charge states of PS ions as well as their ion intensities, which was supported by Llenes et al. and Zhang et al.^17,42 They reported that the metal in the higher oxidation state of FeCl₃, CrCl₃, and CuCl₂ salts involves reduction of Fe(III), Cr(III), and Cu(II) oxidation state to Fe(I), Cr(I), and Cu(I) oxidation state, leading to a slower cationization process to form adduct-PS cations in the gas phase as compared to that with metals in the +1 oxidation state.

Fig. 3 Mass spectra of PS650k obtained with a DCTB matrix using various cationizing reagents with the metal in the +2 oxidation state. The cationizing reagents used were (a) CuBr₂, (b) CuCl₂, and (c) Cu(OAc)₂. The optimal ratio of PS : matrix : cationizing reagent was 3 : 200 : 1 (v/v/v).

The analysis of PS650k with various transition metal salts revealed distinct trends in charge state distributions and signal intensities. Specifically, the order of PS ion signal intensities using different transition metal salts is CuCl₂ < CuBr₂ < Cu(OAc)₂ < CuCl and AgNO₃ < CuBr < CuI < AgTFA. We plotted data based on observations of higher charge states and found that the order of data aligns directly with the lattice energies of the copper salts CuI (948 kJ mol⁻¹), CuBr (969 kJ mol⁻¹), and CuCl (992 kJ mol⁻¹)⁴³ and inversely with their molecular weights and corresponding anion radii (CuI: I-: 220 pm; CuBr: Br-: 196 pm; CuCl: Cl-: 181 pm). In the analysis of PS650k using different Cu salts, a distinct trend was observed where CuI salt yielded superior ion signals with higher charge states and signal intensities as compared to CuBr. When comparing mass spectra obtained with CuBr salt to that with CuCl salt, CuBr yielded a higher charged state PS signal than the CuCl salt. This variation can be attributed to the difference in lattice energies of these salts: CuI (948 kJ mol⁻¹), CuBr (969 kJ mol⁻¹), and CuCl (992 kJ mol⁻¹).⁴³ Higher lattice energies require more energy to dissociate the ionic bonds, thereby affecting the effectiveness of cationization and the formation of PS-cation adducts. Moreover, Cu salts in the +2 oxidation state, such as CuCl₂ (2774 kJ mol⁻¹) and CuBr₂ (2715 kJ mol⁻¹),⁴³ typically produce ions with lower charge states due to their higher lattice energies than the Cu(I) salts and do not directly form adducts with PS. Instead, a favorable reaction involves the reduction of Cu(II) to Cu(I), enabling the formation of lower charge state PS–Cu⁺ adducts. The reaction dynamics suggest that more efficient PS–cation adducts formed directly in a one-step process using Cu(I), transferring more cations to PS neutral and cationized PS to form higher charge state ions with higher signal intensities. In contrast, the two-step process involving Cu(II) salts, typically leads to lower charge state PS ions. Hoberg et al.²⁴ previously investigated the effect of varying lattice energies of halide salts on the ion abundance of poly(methyl methacrylate) (PMMA) samples, finding that salts with lower lattice energies produce stronger PMMA signals. This phenomenon is attributed to the effect of salts with higher lattice energies on the interaction between the salt and analyte within the crystal matrix, which may result in fewer cations being released during ionization, affecting cation attachment to PS.

In the analysis of PS650k, low-intense triply charged dimer peaks were also observed within certain spectra, which led to the examination of mass spectra encompassing these dimer peaks (data not provided). The dimer peaks appeared typically with weak signal intensities, their S/N ratios consistently fell below the threshold of 3–5, and lacked reproducibility across various salts, thus impeding a robust quantitative evaluation of the contribution of multiply charged multimer peaks to monomer peaks signal intensities. Additionally, the settings of ion trapping and collision cooling parameters were optimized primarily for monomer ions, not for multimer ions, which likely exacerbated the inefficacy in cooling and subsequent detection of multimer ions with higher kinetic energies. Given these observations, coupled with the research objective to elucidate how specific reagents influence charge state distributions favoring lower states, the low-intensity multimer peaks were considered to be of marginal significance. Consequently, the analysis of multimer ions in subsequent higher molecular weight samples such as PS900k and PS2M was neglected, anticipating similar challenges in detecting and quantifying these species.

Mass spectra of ∼946 kDa and ∼2 MDa polystyrene ions with Ag and Cu salts

In this study, we only adopted Ag and Cu salts to acquire mass spectra of PS900k (∼946 kDa) and PS2M (∼2 MDa) ions without Cs salt. Because poor singly charged ion signal (S/N ∼ 3) was observed with alkali metal CsCl salt while analyzing PS900k ions as shown in Fig. S4.† Mass spectra of the PS900k sample analyzed with two Ag salts and six Cu salts are shown in Fig. S5.† With AgNO₃ salt (Fig. S5a†) and AgTFA salt (Fig. S5b†), ions with charge states up to +4 and +5 were observed. With CuBr salt (Fig. S5c†) and CuCl salt (Fig. S5d†), ions with charge states up to +4 were observed. With CuBr₂ salt and CuCl₂ salt, ions with charge states up to +3 were observed. With CuI salt (Fig. S5e†), ions with charge states up to +5 were observed. With Cu(OAc)₂ salt (Fig. S5f†), ions with charge states up to +4 were observed. The metal oxidation state of CuBr₂ (Fig. S5g†) and of CuCl₂ (Fig. S5h†) was Cu(II) and the metal oxidation state of CuBr and of CuCl was Cu(I), thereby the charge state of ions reduced from +4 (with CuBr and CuCl) to +3 (with CuBr₂ and CuCl₂), which is consistent with the effect of the metal oxidation state mentioned in the previous section. The charge state distributions, signal intensities, and S/N ratios of PS900k with eight salts are provided in Fig. 4. It is noted that the doubly charged 900k ions showed the highest signal intensities with all salts, except for the CuBr₂ salt. The CuBr₂ salt demonstrated the capability to cationize PS900k ions up to three charge states only, with particularly intense signals observed for singly charged ions. For other salts, the doubly charged ion signal was the most intense and the S/N ratios for the most intense peaks of PS900k with most of the salts exceeded 50. This observation suggests that the ionization efficiency of PS900k with these salts could significantly be enhanced as compared to prior findings where cationization of PS with a molecular weight of 1.1 million daltons (close to PS900k) using Cs salts and the DCTB matrix resulted in signals with S/N < 5.¹¹ The overall performance of CuCl salts closely resembled that of AgNO₃ salts, as evidenced by similar numbers of charge states and S/N ratios in mass spectra across all charge states. Our findings highlight the significant impact of three types of salt (Ag, Cu and Cs) on the mass analysis of megadalton polystyrenes, resulting in significantly improved S/N ratios with different Ag and Cu salts but not with Cs salt.

Fig. 4 Charge state distributions of PS900k ions obtained with a DCTB matrix using various cationizing reagents including AgTFA, AgNO₃, CuI, CuBr, CuCl, CuBr₂, CuCl₂, and Cu(OAc)₂. The optimal ratio of PS : matrix : cationizing reagent was 3 : 200 : 1 (v/v/v).

Furthermore, we measured the PS samples with molecular weights up to two million (PS2M). Mass spectra and the charge distribution of PS2M ions with five different salts are provided in Fig. 5 and S6,† respectively. It was found that CuCl, CuCl₂, and Cu(OAc)₂ showed good ion signal intensities as those with Ag salts. However, several Cu salts, including CuI, CuBr, and CuBr₂, exhibited very weak ion signals (data not shown), making them unsuitable for high-mass PS analysis above 1 MDa. With AgTFA salt, ions with charge states from +2 to +5 were observed. Despite outperforming other salts and demonstrating a signal with S/N > 100 for +2 to +4 charge states, AgTFA salt did not give a singly charged ion signal. In contrast, AgNO₃ salt was capable of generating a broad range of charge states from +1 to +5, thereby offering a more complete charge state distribution, including the singly charged ions. CuCl salt also produced ions with charge states from +2 to +5. However, although the range of charge states for PS2M with CuCl salt is similar to that observed with AgTFA salt, the signal intensities at each charge state were reduced by approximately threefold. In comparison to CuCl salt, CuCl₂ and Cu(OAc)₂ salts yielded PS2M ions with charge states from +1 to +4 and the +1 and +2 ion signals obtained with the Cu(OAc)₂ salt were comparable to those with AgNO₃ salt. The AgNO₃ or Cu(OAc)₂ salts were more efficient than the AgTFA salt in producing +1 ions with PS mass up to 2 MDa, which was not observed in previous studies.^8,11 The obtained results of the 2 MDa polymer using our method are significantly better than those from the previous studies of the 2 MDa polymer using MALDI TOF-MS⁹ and ESI CDMS⁴⁴ techniques. With MALDI TOF-MS, PS samples were analyzed with the use of an STJ cryodetector to achieve UHMW PS analysis in this mass range.⁹ However, the adoption or commercialization of this detector comes with significant cost, which is very expensive as compared with MALDI LIT-MS. The other study employed the CDMS technique to analyze a poly(ethylene oxide) sample with a mass of 2 MDa, revealing charge molecular mass distributions of the polymers at 2.7 MDa.⁴⁴ However, the full width at half maximum (FWHM, Δm) of the mass distribution histogram exceeded 4 MDa, which is about a factor of 40 higher than that obtained with our method, indicating the superiority of our approach in the mass range of 100 kDa to 2 MDa.

Fig. 5 Charge state distributions of PS2M ions obtained with a DCTB matrix using different cationizing reagents including AgTFA, AgNO₃, CuCl, CuCl₂, and Cu(OAc)₂. The optimal ratio of PS : matrix : cationizing reagent was 3 : 200 : 1 (v/v/v).

From the above results, it was found that when PS mass increased from 946 kDa to 2 MDa, the charge states of PS2M ions increased one more charge with AgNO₃ salt and the signal intensities of PS2M ions with +3, +4, and +5 charge states increased with AgTFA salt. The charge state of PS2M ions increased one more charge with CuCl salt and CuCl₂ salt, and the signal intensities of PS2M ions with +2, +3, and +4 charge states increased with Cu(OAc)₂ salt as the PS mass increased from 946 kDa to 2 MDa. It was also observed that the charge states of PS210k (240 kDa), PS650k (606 kDa), and PS900k (946 kDa) ions were up to +2, +3, and +4, respectively with AgNO₃ salt, as shown in Fig. 2. Hence, higher PS mass can gain more charges from transition metal salts, which is because high-mass PS molecules can accommodate more charges with less Coulomb repulsion.⁴⁵ In addition, larger PS molecules could collide with metal cations in the gas phase more frequently and have a faster intramolecular energy redistribution rate due to their larger physical size and more rotational and vibrational modes to form stable polystyrene-metal cations.²⁰

From the mass analysis of various UHMW PS samples above, employing different transition metal cationization salts, Cu salts stand out as a promising option, marking their first use in UHMW PS analysis. One significant advantage of Cu salts over Ag salts is their ability to modulate charge states by varying the lattice energies. Moreover, the alkali metal salt CsCl used in this study, which was previously demonstrated to be superior to Ag salts for analyzing high-mass polymers with the DCTB matrix,¹¹ demonstrates less efficacy than any transition metal salts used in the present study. Table 2 summarizes various cationization salts for mass analysis of UHMW PS samples utilizing the DCTB matrix. The cationization salts are classified based on their +1 and +2 oxidation states. The choice of cationization salts depends on the polymer molecular weight. This information provided in Table 2 serves as a guideline for selecting optimal cationization salts for various UHMW PS samples with the DCTB matrix and might be useful for researchers and analysts in polymer fields. We also found that most Cu salts in the +1 oxidation state (CuBr, CuI, and CuCl) failed to generate high-mass PS ions exceeding 1 MDa. Moreover, salts in the +2 oxidation state could successfully generate ionsfor PS of molecular weight, M_w > 1 MDa.

Table 2 Ionization efficiency of Cu and Ag salts for UHMW PS samples across charge states in DCTB and RA matrices

PS sample	Salt	+1	+2	+3	+4	+5
DCTB
PS650K	AgTFA	56.5	90.4	38.3	14.4	—
	AgNO3	75.4	49.4	10.9	0	—
	CuI	73	74.2	17.6	5.3	—
	CuBr	82.4	68.4	15.3	1	—
	CuCl	92.5	61.1	6.4	—	—
	CuBr2	90	46.7	4.2	—	—
	CuCl2	82.6	50.2	7	—	—
	Cu(OAc)2	88	108.2	16.8	—	—
PS900K	AgTFA	35.2	104.6	44.3	17.3	6.4
	AgNO3	29.5	81.5	37.3	13.1	—
	CuI	25.7	67.3	32.2	15.7	7.1
	CuBr	30.1	35.4	22.1	2.2	—
	CuCl	36.9	81.9	50.2	14.5	—
	CuBr2	35.2	27.6	7.3	—	—
	CuCl2	45.2	55.1	17.9	—	—
	Cu(OAc)2	26.7	49.6	20.8	5	—
PS2M	AgTFA	—	97.9	107.6	76.1	36.9
	AgNO3	11.1	78.7	62.5	34.8	6.4
	CuCl	—	25.1	25.9	12.6	1.2
	CuCl2	2.2	20	17.5	8.1	—
	Cu(OAc)2	10.7	70.1	32.8	12.2	—
RA
PS650K	AgTFA	49.6	113.6	29	13.5	—
	CuCl	81.2	74.1	12.8	—	—
	CuCl2	48.1	67.8	8.4	—	—
	Cu(OAc)2	26.2	53	7.7	—	—
PS900K	AgTFA	13.3	48.8	11	5.7	—
	CuCl	24.3	76.1	13	4.7	—
	CuCl2	18.8	38.7	6.8	3.3	—
	Cu(OAc)2	16.2	26.8	4.3	2.3	—
PS2M	AgTFA	—	23.9	11.7	2.5	1.6
	CuCl	—	10.7	4.1	2.3	—
	CuCl2	—	3.6	2.3	1.4	—
	Cu(OAc)2	—	6.0	4.5	—	—

---

Analysis of UHMW PS using the RA matrix and copper salts

We also performed MALDI LIT-MS analysis on PS650k, PS900k, and PS2M using the RA matrix to assess the compatibility of copper salts with an alternative matrix suitable for UHMW PS analysis. Three copper salts, CuCl, CuCl₂, and Cu(OAc)₂, were chosen for analyzing the three PS samples due to their proven compatibility with the DCTB matrix. Moreover, mass spectra of the three PS samples using the RA matrix in combination with AgTFA salt were used as reference spectra for comparison. This facilitated a comparison of the performance of Cu salts to Ag salts when using the RA matrix. The acquired mass spectra for PS650k, PS900k, and PS2M using the RA matrix with AgTFA salt and Cu salts are presented in Fig. S7, S8, and S9† respectively. Additionally, Fig. 6 illustrates a bar plot depicting the charge distribution, providing insights into the charge state distribution of different PS samples when using the RA matrix with Cu salts.

Fig. 6 Charge state distributions of (a) PS650k, (b) PS900k and (c) PS2M ions obtained with a RA matrix using three different copper cationizing reagents AgTFA, CuCl, CuCl₂, and Cu(OAc)₂. The optimal ratio of PS : matrix : cationizing reagent was 3 : 200 : 1 (v/v/v).

As shown in Fig. S7,† the signal intensity obtained with the RA matrix using AgTFA is comparable to that obtained with the DCTB matrix, though with a disappearance of the +5 charge state when using the RA matrix. The charge distribution of PS650k ions obtained with RA matrices and Cu salts (Fig. 6a) was very similar to the PS signal obtained using DCTB matrices with similar Cu salts. However, using RA matrices with Cu salts resulted in an overall reduction in signal intensities ranging from 20% to 30%. Similarly, when using Cu salts, as well as AgTFA salt, the signal intensities for PS900k (Fig. 6b and S8†) reduced by 40–50%. Further PS2M analysis using the RA matrix (Fig. S9†) resulted in PS signals up to +5 charge states without the singly charged ion signal and a significant reduction in both signal intensities. In addition, only AgTFA and CuCl salts could produce a signal with a S/N ratio greater than 10 (Fig. 6c). Notably, Li et al.⁸ reported a similar observation of the absence of singly charged ions when analyzing PS with a mass above 1 MDa (i.e., PS1.5M) using the Ag salt and RA matrix, which is consistent with our findings. The mass analysis of PS using the RA matrix confirmed that the RA matrix is suitable for PS samples with a mass less than 1 MDa, regardless of whether Ag or Cu salt is used. The analysis of UHMW PS samples using the RA matrix and Cu salt suggests that Cu salts are also suitable with the RA matrix in analyzing UHMW PS samples. However, it is important to note that the DCTB matrix outperformed the RA matrix by providing higher signal intensities and extending the m/z limit of the MALDI MS system up to 2 million Th with selected Ag and Cu salts.

Analysis of the UHMW PS mixture using silver and copper salts

To demonstrate the benefits of Cu(II) salts over Ag salts for a simplified mass analysis of high-mass polymers, we analyzed a mixture of two polymers, PS650k and PS900k. This example helps illustrate how the use of Cu salts can reduce complexity in mass spectra, particularly by managing the generation of multiply charged ions, which often complicates interpretation. As we have previously mentioned the formation of multiply charged ions from Ag and Cu(I) salts may have an adverse effect on mass analysis, especially when the spectra with different charge states of distinct monomers are overlapped. As shown in Fig. 7, mass spectra of the mixture of PS650k and PS900k employing the DCTB matrix with the AgTFA salt show the overlapping of peaks at m/z ranging from 290 kTh to 330 kTh. Mass spectra obtained with AgTFA salt show that the triply charged PS900k peak almost coincides with the doubly charged PS650k peak (red peak in the inset of Fig. 7a). However, with CuCl₂ salt the overlapping of peaks was not observed (black peak in the inset of Fig. 7a). The overlapping of two or more peaks of different polymer very often complicated the identification of two or more polymers in a mixture and could lead to an incorrect interpretation of mass distribution. The challenges in mass distribution analysis using AgTFA thereby become more intricate when analyzing UHMW polymer mixtures. Additionally, the signal intensity for singly charged peaks of PS900k was initially poor with Ag salts but improved significantly when using Cu salts. Copper salts facilitate the generation of lower charge states, enhancing the S/N ratio for singly charged peaks and reducing the formation of higher charge states in UHMW PS mass analysis. This improvement helps prevent the overlapping of peaks from polystyrene samples of different masses in a mixture. The advantage of using Cu salts stems from its capacity to modulate charge states through the use of salts derived from various oxidation states of copper. This underscores the significance of utilizing Cu salts, preliminarily generating low charge state ions and offering a more suitable approach for the characterization of UHMW polymer mixtures.

Fig. 7 Mass spectra of the mixture of PS650k and PS900k obtained with a DCTB matrix using (a) AgTFA, and (b) CuCl₂ cationization reagents. The inset in (a) shows a zoomed-out view of peaks ranging from m/z 275kTh to 360kTh. The peak highlighted in red is with AgTFA, while in black is with CuCl₂ salt.

Table 2 contains the S/N ratios of the PS signals across different charge states with DCTB and RA matrices, emphasizing the ionization efficiency of Cu and Ag salts for UHMW PS samples. In the DCTB matrix, Cu salts—both Cu(I) and Cu(II)—show higher S/N ratios at lower charge states (+1 and +2), suggesting strong ionization efficiency which is beneficial for clear spectral output in complex polymer analyses. However, as charge states increase, there is a noticeable decline in these ratios, indicating reduced stability of higher charged ions. Conversely, Ag salts in the DCTB matrix display a progressive increase in S/N ratios as charge states increase, hinting at their capability to effectively stabilize and ionize higher charged states as compared to the use of Cu salts. In the RA matrix, the pattern shifts slightly: Cu salts still exhibit good ionization efficiency at lower charge states, but with generally lower S/N ratios compared to the use of the DCTB matrix, suggesting a matrix-specific interaction that influences ionization dynamics. Ag salts, particularly AgTFA, show stronger ionization capabilities in higher charge states within the RA matrix. These observations highlight the importance of selecting the appropriate cationizing reagents and matrices based on the complexity of the polymer systems.

Conclusion

This study demonstrates the advantages of copper (Cu) salts over silver (Ag) and cesium (Cs) salts in the matrix-assisted laser desorption/ionization (MALDI) analysis of ultra-high molecular weight (UHMW) polystyrene, achieving detection up to two million daltons. The study highlights the distinctive advantages of Cu salts, particularly in their ability to effectively modulate ion charge states, which enhances analytical precision and flexibility. Cu(I) salts, which generate charge states up to +5, enable numerous independent mass measurements of a molecule, which can be averaged to enhance mass determination accuracy. Conversely, Cu(II) salts, which produce lower charge states up to +3, offer advantages in minimizing charge state overlap in complex polymer mixtures, thereby facilitating clearer and more interpretable mass spectra. This targeted control over ion charge states with both Cu(I) and Cu(II) salts greatly simplifies the analysis of both individual and mixed polymer samples, enhancing methodological flexibility. Ag salts can be advantageous for detailed mass analysis of simpler polymer structures, as they generate high-intensity, multiply charged ion signals for UHMW PS samples. However, their tendency to produce these ions without specific charge state control can complicate spectral interpretation, especially in polymer mixtures. The Cu salts are compatible with DCTB and RA matrices in analyzing the UHMW PS, where DCTB outperformed the RA matrix. This study emphasizes the practical applications of Cu salts in advanced MALDI polymer analysis, showcasing their cost-effectiveness and potential in mass analysis across a wide range of UHMW polymers^2,46 and polymer dendrimers.¹⁰

Data availability

Additional data are made available in ESI† tables of this manuscript.

Conflicts of interest

The authors have no conflicts of interest to declare.

Acknowledgements

This work was partially supported by the National Science and Technology Council (NSTC) of Taiwan through grant numbers NSTC 113-2112-M-259-002 (W. P. P.), 112-2112-M-259-015 (W. P. P.), 113-2112-M-259-006 (W. P. P.), 114-2923-M-259-001-MY3 (W. P. P.), and 112-2811-M-259-005 (A. A. P.), and National Dong Hwa University.

References

1. V. H. Dao, N. R. Cameron and K. Saito, Macromolecules, 2019, 52, 7613–7624.
2. A. Reyhani, S. Allison-Logan, H. Ranji-Burachaloo, T. G. McKenzie, G. Bryant and G. G. Qiao, J. Polym. Sci., Part A: Polym. Chem., 2019, 57, 1922–1930.
3. Y. Wang, Q. Wang and X. Pan, Cell Rep. Phys. Sci., 2020, 1, 100073.
4. S. Chhetri and H. Bougherara, Composites, Part A, 2021, 140, 106146.
5. N. Aust, J. Biochem. Biophys. Methods, 2003, 56, 323–334.
6. D. Lee, Y. Zhu, B. Chae and E. J. Tocce, J. Chromatogr. A, 2021, 1659, 462640.
7. J. Wang, H. Huang and X. Huang, J. Appl. Polym. Sci., 2016, 133(31), 43748.
8. D. C. Schriemer and L. Li, Anal. Chem., 1996, 68, 2721–2725.
9. A. A. Aksenov and M. E. Bier, J. Am. Soc. Mass Spectrom., 2008, 19, 219–230.
10. H. J. Räder, T.-T.-T. Nguyen and K. Müllen, Macromolecules, 2014, 47, 1240–1248.
11. S. J. Gabriel, R. F. Steinhoff, M. Pabst, C. Schwarzinger, R. Zenobi, U. Panne and S. M. Weidner, Rapid Commun. Mass Spectrom., 2015, 29, 1039–1046.
12. A. A. Patil, Z.-X. Liu, Y.-P. Chiu, T. K. L. Lại, S.-W. Chou, C.-Y. Cheng, W.-M. Su, H.-T. Liao, J. B. A. Agcaoili and W.-P. Peng, Talanta, 2023, 259, 124555.
13. A. M. Belu, J. M. DeSimone, R. W. Linton, G. W. Lange and R. M. Friedman, J. Am. Soc. Mass Spectrom., 1996, 7, 11–24.
14. S. J. Wetzel, C. M. Guttman and J. E. Girard, Int. J. Mass Spectrom., 2004, 238, 215–225.
15. S. F. Macha, P. A. Limbach and P. J. Savickas, J. Am. Soc. Mass Spectrom., 2000, 11, 731–737.
16. T. Habumugisha, Z. Zhang, J. C. Ndayishimiye, F. Nkinahamira, A. Kayiranga, E. Cyubahiro, A. Rehman, C. Yan and X. Zhang, Anal. Methods, 2022, 14, 763–772.
17. C. F. Llenes, R. M. O'Malley and R. J. Cotter, Rapid Commun. Mass Spectrom., 1992, 6, 564–570.
18. P. O. Danis, D. E. Karr, Y. Xiong and K. G. Owens, Rapid Commun. Mass Spectrom., 1996, 10, 862–868.
19. M. J. Deery, K. R. Jennings, C. B. Jasieczek, D. M. Haddleton, A. T. Jackson, H. T. Yates and J. H. Scrivens, Rapid Commun. Mass Spectrom., 1997, 11, 57–62.
20. H. Rashidezadeh and B. Guo, J. Am. Soc. Mass Spectrom., 1998, 9, 724–730.
21. H. Chen, M. He, Z.-K. Chen and C.-Z. Zhou, J. Am. Soc. Mass Spectrom., 2005, 16, 1695–1701.
22. G. P. Zeegers, R. F. Steinhoff, S. M. Weidner and R. Zenobi, Int. J. Mass spectrom., 2017, 416, 80–89.
23. T. Yalcin, D. C. Schriemer and L. Liang, J. Am. Soc. Mass Spectrom., 1997, 8, 1220–1229.
24. A.-M. Hoberg, D. M. Haddleton, P. J. Derrick, A. T. Jackson and J. H. Scrivens, Eur. J. Mass Spectrom., 1998, 4, 435–440.
25. Y. Wang, H. Rashidzadeh and B. Guo, J. Am. Soc. Mass Spectrom., 2000, 11, 639–643.
26. A. R. Hortal, P. Hurtado, B. Martínez-Haya, A. Arregui and L. Bañares, J. Phys. Chem. B, 2008, 112, 8530–8535.
27. G. Montaudo, M. S. Montaudo, C. Puglisi and F. Samperi, Rapid Commun. Mass Spectrom., 1995, 9, 453–460.
28. F. J. Cox, M. V. Johnston and A. Dasgupta, J. Am. Soc. Mass Spectrom., 2003, 14, 648–657.
29. H.-L. Hsieh, C.-Y. Cheng, L. Yi-Kun, C.-H. Yang, F. Liang-Chun, Y.-h. Tseng, S.-W. Chou and Acromass Technologies Inc, US Pat., US10984999B2, 2021.
30. V. K. Sharma and F. J. Millero, Geochim. Cosmochim. Acta, 1989, 53, 2269–2276.
31. M. Fondo, J. Chem. Educ., 2023, 100, 4791–4795.
32. A. A. Patil, C. K. Chiang, C. H. Wen and W. P. Peng, Anal. Chim. Acta, 2018, 1031, 128–133.
33. A. A. Patil, T. K. Ly Lại, C.-W. Wang, G.-F. Chen, B.-X. Du, C.-K. Chiang and W.-P. Peng, Int. J. Mass Spectrom., 2021, 462, 116539.
34. K. Yun, I. Jalaludin, S. Jung, K.-S. Jang and J. Kim, Analyst, 2022, 147, 505–515.
35. C. Lee, I. C. Lu, H. C. Hsu, H.-Y. Lin, S.-P. Liang, Y.-T. Lee and C.-K. Ni, J. Am. Soc. Mass Spectrom., 2016, 27, 1491–1498.
36. A.-M. Hoberg, D. M. Haddleton, P. J. Derrick and J. H. Scrivens, Eur. J. Mass Spectrom., 1997, 3, 471–473.
37. R. Knochenmuss, E. Lehmann and R. Zenobi, Eur. J. Mass Spectrom., 1998, 4, 421–427.
38. S. D. Hanton, K. G. Owens, C. Chavez-Eng, A.-M. Hoberg and P. J. Derrick, Eur. J. Mass Spectrom., 2005, 11, 23–29.
39. W. J. Erb, S. D. Hanton and K. G. Owens, Rapid Commun. Mass Spectrom., 2006, 20, 2165–2169.
40. J. B. Metternich, M. F. Czar, M. F. Mirabelli, G. L. Bartolomeo, K. C. M. Zouboulis and R. Zenobi, J. Am. Soc. Mass Spectrom., 2019, 30, 2392–2397.
41. Z. Wang, Q. Zhang, H. Shen, P. Yang and X. Zhou, Front. Chem., 2021, 9, 698297.
42. J. Zhang, V. Frankevich, R. Knochenmuss, S. D. Friess and R. Zenobi, J. Am. Soc. Mass Spectrom., 2003, 14, 42–50.
43. Handbook of Chemistry and Physics, ed. W. M. Haynes, CRC Press Taylor & Francis Group, 95th edn, 2014–2015.
44. R. Antoine, Rapid Commun. Mass Spectrom., 2020, 34, e8539.
45. K. Vékey, Mass Spectrom. Rev., 1995, 14, 195–225.
46. R. Li and Z. An, Angew. Chem., Int. Ed., 2020, 59, 22258–22264.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ay02077d

‡ These two authors contributed equally to this work.

---