Multidimensional Mass Spectrometry Coupled with Separation by

Dec 7, 2015 - The product supposedly contained an average of 20 EO units (20 ...... to study composition and base hydrolysis of polysorbate 20 and 80 ...
1 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/ac

Multidimensional Mass Spectrometry Coupled with Separation by Polarity or Shape for the Characterization of Sugar-Based Nonionic Surfactants Bryan C. Katzenmeyer,† Shayna F. Hague,‡ and Chrys Wesdemiotis* Department of Chemistry, The University of Akron, Akron, Ohio 44325, United States S Supporting Information *

ABSTRACT: Mass spectrometry (MS) and tandem mass spectrometry (MS/MS) were interfaced with ultra-performance liquid chromatography (UPLC) and ion mobility (IM) separation to characterize a complex nonionic surfactant, consisting of a methylated glucose core (glucam) conjugated with poly(ethylene oxide) (PEOn) branches that were partially esterified with stearic acid to form ethoxylated glucam (PEOnglucam) stearates. Reverse-phase LC-MS afforded fast separation according to polarity into five major fractions. Accurate mass measurements of the ions in the mass spectra extracted from these fractions enabled conclusive identification of six components in the surfactant, including PEOn-glucam mono-, di-, and tristearates as well as free and esterified PEOn as byproducts. MS/MS experiments provided corroborating evidence for the fatty acid content in each fraction based on the number of stearic acid losses observed. With IM-MS, the total surfactant ions were separated according to charge and shape into four distinct bands. Extracted mass spectra confirmed the presence of two disaccharide stearates in the surfactant, which were undetectable by LC-MS. PEOn-glucam tristearates were, however, not observed upon IM-MS. Hence, LC-MS and IM-MS unveiled complementary compositional insight. With each method, certain components were particularly well separated from other ingredients (by either polarity or shape), to be detected with confidence. Consequently, combined LC-MS and IM-MS offer a superior approach for the characterization of surfactants and other amphiphilic polymers and for the differentiation of similarly composed amphiphilic blends. It is finally noteworthy that NH4+ charges minimized chemical noise in MS mode and Li+ charges maximized the fragmentation efficiency in MS/MS mode.

S

reveals specific compositional and structural insight. LC-MS has been used to analyze a variety of surfactants, including polysorbates,11−14 alkyl ethoxylates,7,8,15−18 alkyl polysaccharides,19,20 and copolymers.21,22 More recently, Scrivens and coworkers24 combined atmospheric solids analysis probe (ASAP)23 ionization and matrix-assisted laser desorption/ ionization (MALDI) with ion mobility mass spectrometry (IM-MS) and tandem mass spectrometry (MS/MS) to determine the percentage content of individual esters in polysorbates and to categorize their major, minor, and trace ingredients. Similar results have been obtained by interfacing electrospray ionization (ESI) with LC-MS, IM-MS, and MS/ MS.25 Synthetic macromolecules have traditionally been fractionized by size exclusion chromatography (SEC).26,27 The typical solvents used in SEC, however, are not optimal for online coupling with ESI.28 On the other hand, the polar mobile phases generally utilized in reverse-phase liquid chromatog-

ugar-based nonionic surfactants are amphiphilic polymers with low toxicity and rapid biodegradability that are added as emulsifiers, detergents, dispersants, and/or foaming agents to a large number of commercial and industrial products.1 Biomedical applications of these materials are also being exploited for the encapsulation and delivery of therapeutic and diagnostic drugs.2 Nonionic surfactants are typically blends of homologous (macro)molecules and may contain additional chemical species, formed during their formulation.3,4 Given this complexity, a palette of methods has been developed to separate and characterize these compounds.5 High-performance liquid chromatography (HPLC or LC) with UV−vis detection has been used widely to identify the components of nonionic surfactants.6−8 Unfortunately, most modern surfactants do not possess UV−vis absorbing chromophores due to increased environmental concerns. Although refractive index (RI)9 and evaporative light scattering (ELS)10 detectors can be used in such cases, these detection systems (like UV−vis detectors) lack the specificity needed to unravel the complexity of nonionic surfactant blends. These problems are largely bypassed by coupling LC to mass spectrometry (MS) which © XXXX American Chemical Society

Received: September 6, 2015 Accepted: December 7, 2015

A

DOI: 10.1021/acs.analchem.5b03400 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

final concentration of 10 μM in methanol/water (50:50 v/v). A 5 μM [Glu1]-Fibrinopeptide B solution in water/methanol/ formic acid (67:33:0.1 v/v/v) was used as an external standard for accurate mass measurements. Liquid Chromatography. Reverse-phase LC was carried out using an Aquity UPLC system (Waters Corporation, Milford, MA). Separation was achieved on a Waters BEH C18 column (2.1 × 50 mm, 1.7 μm particle size) maintained at 60 °C. Two mobile phases (A and B) were employed; mobile phase A consisted of a 2.50 mM solution of ammonium acetate in water/methanol (97:3 v/v; pH not adjusted), and mobile phase B was 100% methanol. Surfactant components were separated by combined gradient and isocratic elution at a flow rate of 400 μL min−1 with the percentage of B varied as follows: linear increase from 0% to 40% over 2 min, linear increase to 60% over 1 min, and linear increase to 100% over 4 min and then 100% B for an additional 4 min. Sample volumes of 5−10 μL were injected into the UPLC column. A 2.50 mM solution of ammonium acetate in methanol was used as a postcolumn additive during LC-MS. For LC-MS/MS analysis, a 2.50 mM lithium acetate replaced ammonium acetate in mobile phase A and the postcolumn additive. Mass Spectrometry. Accurate mass measurements were acquired using a Synapt HDMS quadrupole/time-of-flight (Q/ TOF) mass spectrometer (Waters Corporation, Milford, MA). The triwave region of this instrument, located between the Q and TOF mass analyzers, contains three confined regions in the order trap cell (closest to Q), ion mobility (IM) cell, and transfer cell (closest to TOF). The trap and transfer cells are pressurized with argon (Ar gas flow 1.5 mL min−1) and either one can be used for conventional MS/MS experiments via collisionally activated dissociation (CAD). The IM cell is used in IM-MS experiments (see below). The Synapt utilizes a z-spray source with two inlets, one serving to introduce the analyte and the other acting as a spray for the lock mass reference standard. Calibration was based upon the [Glu1]-Fibrinopeptide B doubly charged ion at m/z 785.8426 ([M + 2H]2+). This procedure allows mass measurements within ≤10 ppm of the calculated mass.48 Experiments were performed using the following parameters: ESI capillary voltage, 3.5 kV; sample cone voltage, 25 V; extraction cone voltage, 1.0 V; source temperature, 90 °C; desolvation temperature, 250 °C; cone gas flow, 21 L h−1; desolvation gas flow, 650 L h−1 (N2). Mass spectra were acquired with the TOF analyzer (Q in rf-only mode). For MS/ MS, Q was adjusted to transmit only ions of the desired m/z ratio; these were then subjected to collisionally activated dissociation (CAD) with argon gas in the trap cell, and the resulting fragments were mass-analyzed by the TOF analyzer. Collision energies were set by varying the trap cell potential between 53 and 79 V, while the transfer cell potential was kept at 2 V. Ion Mobility Mass Spectrometry. ESI-MS parameters were the same as those above, and IM parameters were set as follows: IM gas flow, 14.0 mL min−1 (N2); traveling wave velocity, 500 m s−1; traveling wave height, 10 V. Samples were injected directly into the ESI source at a flow rate of 10 μL min−1, using a syringe pump; they were prepared by dissolving 2.5 mg of surfactant in methanol/water (50/50 v/v), diluting with this solvent to a final concentration of 0.025 mg mL−1, and mixing the resulting solution with a few droplets of aqueous ammonium acetate (100 mg mL−1) in the ratio of 100:1 (v/v).

raphy are most appropriate for introduction into the ESI source. Reverse-phase LC-MS has been successfully applied to the separation and identification of low molecular weight polymer blends, including nonionic surfactants.15,17,18,20,22,25,29−31 Compared to HPLC, the arrival of LC columns with sub-2 μm particles has enabled ultraperformance LC (UPLC),32,33 which offers better separation efficiency resulting in improved resolution, sensitivity, and speed over traditional HPLC. These benefits are further amplified by (UP)LC-MS with a mass spectrometer of higher resolving power so that accurate mass-to-charge ratios (m/z) can be acquired to distinguish isobars. UPLC-MS requires approximately 10−30% of the time needed for HPLC-MS, which minimizes solvent use. An even faster separation is achievable by IM-MS,34−39 which is especially valuable for labile or weakly bound macromolecular systems that would be destroyed or permanently retained with LC and, hence, would remain invisible and unidentifiable unless an alternative separation technique is employed.31,40−47 Both of these methods are employed in this study to characterize a complex, sugar-based nonionic surfactant. The amphiphilic system examined is a star-branched polymer composed of a methylated glucose core (glucam) that was chain-extended with ethylene oxide (EO) to give poly(ethylene oxide) derivatized glucam, PEOn-glucam, whose OH end groups were partially esterified with stearic acid (C18H36O2) to make “poly-ethoxylated glucam stearates.” The product supposedly contained an average of 20 EO units (20 “ethoxylations”) and one-and-half mol of esterified chain ends per mol of surfactant (“sesquistearate”), cf. Figure 1. Here, we

Figure 1. Star-branched nonionic surfactant, consisting of a methylated glucose core (glucam), n ethoxylations on each free saccharide hydroxyl group (n = 5 on average), and a varying number of stearate ester groups (1.5 mol of hydrophobic esters per mol of hydrophilic glucam).

report the first UPLC-MS and MS/MS investigation of a synthetic polymer, concerning the characterization of the described surfactant star. Interactive, reverse-phase (RP) chromatographic mode was used which separates analyte components by their hydrophobicity−hydrophilicity balance.18,25,26 The efficacy of this polarity-based approach is compared to the size and shape sensitive separation effected by IM-MS.24,25,34 The advantages and shortcomings of each method are also discussed briefly.

■

EXPERIMENTAL SECTION Materials. Water, methanol, and ammonium acetate (all of LC-MS grade) were acquired from Fisher Scientific (Pittsburgh, PA). LC-MS grade lithium acetate, formic acid, and [Glu1]Fibrinopeptide B (human) were acquired from Sigma-Aldrich (St. Louis, MO). The nonionic surfactant poly(ethylene oxide) (PEOn)-glucam sesquistearate was obtained from the Lubrizol Corporation (Wickliffe, OH). All chemicals were used as received. For LC-MS and LC-MS/MS analyses, the samples were dissolved in methanol and diluted with water to give a B

DOI: 10.1021/acs.analchem.5b03400 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

■

RESULTS AND DISCUSSION The synthesis of ethoxylated sugar-based surfactants, like the one shown in Figure 1, poses an analytical challenge because the product is generally composed of a distribution of oligomers differing in the degree of poly(ethylene oxide) (PEO or PEOn) polymerization and stearate esterification. In addition, PEO byproducts may be formed that are isobaric with the main polymer constituents (cf. Table S1). The inclusion of the glucam core in PEO or PEO-stearate(s) raises the oligomer mass by 176.068 Da. This increment is isobaric with the mass of four EO units (176.105 Da). Hence, analysis at high mass accuracy is necessary to determine with confidence whether oligomer masses contain the glucam core or not. ESI-MS analysis by direct infusion results in superimposed distributions in several charge states (cf. Figure S1) which complicates spectral analysis and disables a conclusive characterization. To overcome these problems, the addition of a separation dimension to mass analysis was examined in this study. Two approaches were considered, viz., ultra-performance liquid chromatography (UPLC) and ion mobility (IM) spectrometry, which separate preionization in the solution phase and postionization in the gas phase, respectively. Separation and Component Identification by UPLCMS. Combined gradient/isocratic elution and ESI-MS detection of the eluates gave rise to the total ion chromatogram (TIC) shown in Figure 2, in which several well resolved

product as PEOn-glucam stearate (see structure in Figure 3). It is noteworthy that ESI of the oligomers in this component gives rise to singly as well as doubly and triply charged distributions, all cationized by NH4+. Similar LC-MS analysis of the other bands in the TIC chromatogram (Figure 2) led to the compositional assignments summarized in Table 1. The first two fractions, with TIC maxima at 0.41 and 2.74 min, contain PEOn with H− and −OH end groups and a mass distribution of 44n + 18 Da (see Figures S2 and S3). The shorter chains detected at the lowest retention time (0.41 min) are attributed to hydrogen-bonded PEOn aggregates of high polarity; conversely, the PEOn eluting later (2.74 min) is allotted to the longer chains that cannot interact with each other strongly enough to form association products. The elution of hydrogen-bonded aggregates and longer, nonassociated polymer chains at distinct retention times has been previously observed during the SEC fractionation of OHterminated polycarbonates.49 In our experiments, the very short retention time of the PEOn aggregates (near the column void volume peak) must result from their strong hydrogen bonding interactions with the water molecules in the mobile phase. Oligomers carrying one stearate ester are observed next at significantly longer retention times, with the PEOn-glucam stearate eluting earlier than PEOn stearate (TIC maxima at 6.48 and 6.66 min, respectively). This elution order is reasonable, because the sugar moiety increases the hydrophilicity due to higher oxygen content, weakening interactions with the stationary phase and speeding up elution. The corresponding LC-MS spectra are shown in Figures 3 (PEOn-glucam stearate) and S4 (PEOn stearate). Distearates are found in the TIC band between 7.0 and 8.0 min, which shows a maximum at 7.83 min. Single LC-MS scans across this peak reveal that PEOn-glucam distearate (Figure S5) is carried faster through the column than plain PEOn-distearate (Figure S6), as expected, due to the higher polarity of the former compound. Finally, the last band, at 9.66 min, contains the most hydrophobic component, viz., PEOn-glucam tristearate (Figure S7). These results clearly show that LC separation is primarily controlled by the degree of hydrophobicity of the surfactant constituents, with the more hydrophobic species exhibiting longer retention times because of more favorable interactions with the hydrocarbon-based stationary phase. Confirmation of Degree of Esterification by UPLC-MS/ MS. The fatty acid ester content of a particular fraction was further probed by the fragmentation characteristics observed in LC-MS/MS spectra acquired by CAD with argon gas. Collisionally activated polymer ions with ester functionalities can undergo charge-remote 1,5-H rearrangement over the ester

Figure 2. UPLC-MS chromatogram (TIC vs time) of PEOn-glucam sesquistearate. The retention times at the peak maxima are marked.

fractions are clearly discerned. NH4+ ions were used for cationization, as they led to higher ion intensities and less chemical noise compared to Na+ or Li+ adduction. With reverse-phase chromatography, eluates are expected to increase in hydrophobicity as the retention time increases; this expectation is corroborated by the LC-MS spectra extracted from the resolved chromatographic bands. A representative LC-MS spectrum is shown in Figure 3, acquired from the eluate with a retention time of 6.48 min (single scan); the corresponding m/z values identify this

Figure 3. LC-MS spectrum of the eluate with a retention time of 6.48 min from PEOn-glucam sesquistearate; triply (green ◆), doubly (red ■), and singly (blue +) charged ions with the composition [M + xNH4]x+ are observed. Their monoisotopic m/z values indicate that this fraction contains PEOn-glucam monostearate with a mass distribution of 44n + 20 Da (Table 1). The ester may be attached at any of the PEO chains. C

DOI: 10.1021/acs.analchem.5b03400 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry Table 1. PEOn-Glucam Sesquistearate Constituents Identified by UPLC-MS and Accurate Mass Measurement retention time (min) 0.41 2.74 6.48 6.66 7.83 8.04 9.66

measured m/z (charge) 599.378 643.401 754.502 830.627 887.633 1140.919 1019.712

(+2) (+2) (+2) (+1) (+2) (+1) (+2)

chemical formula

calculated m/z

error (ppm)a

assignmentb

C52H106O27(NH4)2 C56H114O29(NH4)2 C71H140O30(NH4)2 C42H82O13NH4 C89H174O31(NH4)2 C62H122O16NH4 C101H198O37(NH4)2

599.380 643.407 754.506 830.621 887.636 1140.908 1019.715

4.1 8.8 5.1 7.9 3.8 9.9 2.9

PEO26 PEO28 PEO23-glucam stearate PEO12 stearate PEO27-glucam distearate PEO13 distearate PEO29-glucam tristearate

Error in measured m/z. bSubscripts indicate the number of EO repeat units. All structures also contain the elements of H2O (H− and −OH end groups) which have been omitted for brevity. a

Figure 4. LC-MS/MS spectra of the [M + 2Li]2+ ions of (a) PEO24-glucam distearate (m/z 898.55) and (b) PEO24-glucam tristearate (m/z 1031.77) from the LC fractions eluting at 7.83 and 9.66 min, respectively. The precursor ions (marked by a red star) were subjected to CAD with Ar at collision cell potentials of (a) 72 and (b) 79 V. The m/z values marked are monoisotopic. Loss of one stearic acid molecule (C18H36O2) decreases the mass by 284 Da and the m/z value by 284 (1+) or 142 (2+) m/z units. The ion at m/z 311 (marked by ∗) corresponds to the stearic acid dioxolanylium fragment (see structure in Scheme S2).

group, if an H atom is available in γ position to the carbonyl moiety (cf. Scheme S1).50,51 From ethoxylated surfactants that are esterified with a fatty acid, this reaction releases the fatty acid as a neutral moiety to yield a fragment ion with an alkene end group (Scheme S1).25 The mass of the neutral loss reveals the identity of the fatty acid and the number of such losses, the degree of esterification. This structurally diagnostic fragmentation proceeds readily in alkali metal cationized oligomers but is suppressed in [M + xNH4]x+ ions which mainly undergo NH3 loss.52 Li+ cationization was employed in this study, as it led to more abundant and clearly discernible fatty acid losses compared to Na+ cationization. In order to promote lithiation of the eluting components in the UPLC-MS/MS experiments, a lithium salt was used in both the mobile phase and a postcolumn additive (vide supra). The major oligomer across a single LC-MS scan was selected as precursor ion for MS/MS. Fragmentation was effected by ramping the collision energy until the dominant fragment peak was approximately twice the height of the precursor peak. The LC fractions with no fatty acid ester content (eluting at 0.41 and 2.74 min) showed only fragment ions consisting of PEOn chains with various terminal groups (mainly H− and −OH), which is indicative for free, dihydroxy-terminated PEOn (see Figures S8 and S9).51 Following these fractions, PEOn-glucam stearate and PEOn stearate (TIC maxima at 6.48 and 6.66 min, respectively) showed a single loss of stearic acid (cf. Figures S10 and S11). The last two major fractions comprised distearates and tristearates (7.83 and 9.66 min, respectively).

The LC-MS/MS spectra of [M + 2Li]2+ ions from PEOnglucam distearate and PEOn-glucam tristearate are shown in Figure 4. They clearly document the elimination of up to two stearic acid molecules from the distearate and up to three stearic molecules from the tristearate, and in both cases, the resulting fragment ions are observed singly as well as doubly charged. Lastly, the LC-MS/MS spectrum of PEOn distearate reveals that up to two stearic acid molecules are eliminated (Figure S12), consistent with the distearate structure of this surfactant constituent. As mentioned above, the loss of stearic acid results from a charge-remote 1,5-hydrogen rearrangement over the fatty ester moiety. A competitive dissociation channel of ethoxylated fatty esters involves charge-induced fragmentation to a dioxolanylium cation with an m/z value characteristic of the fatty acid (cf. Scheme S2);12,24,25 the stearate dioxolanylium fragment is observed at m/z 311 (marked by ∗ in the LC-MS/MS spectra). The relative abundance of this fragment, which provides complementary information about the presence of stearate(s) in the surfactant, increases considerably with the number of such substituents (cf. Figure 4). It is worth noting that both the LC-MS and the LC-MS/MS data confirm that no other fatty acid besides stearic acid was used in the synthesis. If a natural fat were used for PEOnglucam esterification, fatty esters other than stearate (for example, palmitate or linoleate) would have been detected. The peak areas in the UPLC-MS chromatogram of Figure 2 reveal that approximately 69% of the surfactant molecules D

DOI: 10.1021/acs.analchem.5b03400 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

linear structures. The PEOn-glucam sesquistearate components identified in bands 2−4 were also observed by UPLC-MS, except for nonesterified PEOn-glucam which was not detected using LC-MS presumably because of its very small quantity in the analyzed sample (vide infra). The IM-MS mass spectrum extracted from band 1 (Figure 6) indicates the presence of two triply ammoniated disaccharides

contain the glucam core and that 83% of the PEOn and PEOnglucam components are esterified (cf. Table S2). Because of the high degree of esterification, the majority of molecules in this compound are amphiphiles, carrying both hydrophilic and hydrophobic substituents. On the basis of the relative peak areas of the PEOn-glucam mono-, di-, and tristearates, the average number of stearate groups per PEOn-glucam molecule is estimated at approximately 1.6, in good agreement with the quoted 1.5. Separation and Component Identification by IM-MS. IM-MS34−38 and its traveling wave variant (TWIM-MS)38,53,54 may be viewed as postionization chromatographic methods that separate gas-phase ions according to their mass, charge, and shape. Direct infusion of PEOn-glucam sesquistearate into the ESI source, followed by dispersion of the ionized oligomers in the IM chamber and subsequent mass analysis, gave rise to four discrete bands of ions, cf. Figure 5. In the IM chamber, ions

Figure 6. Mass spectrum extracted from band 1 in the IM-MS plot of PEOn-glucam sesquistearate (Figure 5). It includes two triply charged ammoniated distributions, arising from PEOn-diglucam stearate (red ○) and PEOn-diglucam distearate (green ●). The inset shows the structure of PEOn-diglucam. The stearate(s) may be attached to any of the PEO chains. Figure 5. Two-dimensional IM-MS plot (m/z vs drift time) of PEOnglucam sesquistearate, showing four distinct bands (1−4) of mobilityseparated ions with one or more NH4+ charges; band 1 contains triply charged PEOn-diglucam stearate and PEOn-diglucam distearate (cf. Figure 6); band 2 contains doubly charged PEOn-glucam, PEOnglucam stearate, and PEOn-glucam distearate (Figure S13); band 3 contains singly charged PEOn (Figure S14); band 4 contains singly charged PEOn stearate and PEOn distearate (Figure S15).

in this mobility region, corresponding to PEOn-diglucam stearate and PEOn-diglucam distearate and having mass distributions of 44n + 4 Da and 44n + 6 Da, respectively. This result suggests that there was a trace amount of nonmethylated glucose present which reacted with a methylated glucose molecule to give a disaccharide (diglucam) that ultimately was ethoxylated and esterified to give the two distributions shown in Figure 6. The ion intensities in band 1 are consistent with trace quantities for the disaccharides, justifying why they were missed upon LC-MS. Fortunately, their higher charge states and distinct CCSs (vs those of the other surfactant components) place them in a unique mobility region in the IM-MS experiment, allowing for their detection and characterization. The other PEOn-glucam sesquistearate component observed by IM-MS, but not LC-MS, is nonesterified PEOn-glucam (vide supra). Shape sensitive separation (IM-MS) grouped this constituent together with its stearates, where it could be detected despite its small quantity in the analyzed sample because esterification changes the mass (cf. Figure S13), whereas polarity-based separation presumably dispersed PEOn-glucam together with the isobaric, similarly polar, and very easily ionizable linear PEOn, thereby suppressing its detection. On the other hand, PEOn-glucam tristearate was observed by LC-MS, but not by IM-MS; this is attributed to the lower ionization efficiency of the more hydrophobic tristearate, which compromises its detection when it is mixed with more hydrophilic constituents of similar size/ shape.55,56 Obviously, each method, LC-MS and IM-MS, offers unique information not clearly available by the other method and using both enables a superior compositional and structural characterization.

travel within a bath gas under the influence of an electric field and, in this process, they are dispersed according to their collision cross-section (CCS) and charge. CCS represents a measure of the ions’ forward moving area and is a function of ion size (mass) and shape (architecture). Linear, extended structures have higher CCSs and move more slowly through the IM chamber than more compact or cyclic architectures. Meanwhile, higher charge states move faster through the IM chamber due to higher drifting velocities.34−47 The components in each of the separated bands in Figure 5 can be identified by the mass spectra extracted from them; those of bands 2, 3, and 4 are shown in Figures S13, S14, and S15, respectively. On the basis of measured m/z values, the following compositions were deduced: a mixture of doubly ammoniated PEOn-glucam, PEOn-glucam stearate, and PEOnglucam distearate in band 2; singly ammoniated PEOn in band 3; and a mixture of singly ammoniated PEOn stearate and PEOn distearate in band 4. Expectedly, the PEOn stearates (band 4), which contain long alkyl chains that cannot interact with the charge, have longer drift times than nonesterified PEOn (band 3), in which multiple hydrogen bonds between the oxygen sites and the NH4+ charge enable the formation of more compact structures. On the other hand, the glucam-containing species in band 2 (all doubly charged) are not separated well by the degree of esterification, suggesting a much smaller dependence of CCS on fatty ester content for star-branched as compared to E

DOI: 10.1021/acs.analchem.5b03400 Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

■

■

CONCLUSIONS Mass spectrometry is employed in many fields due to its exceptional sensitivity, low detection limits, and high speed of analysis. Despite these advantages, mass spectrometry faces limitations when analyzing complex mixtures, such as polymer blends, due to discrimination effects in the ionization and/or detection steps and the inability to distinguish isomeric and often closely isobaric analyte components. These issues can be circumvented by coupling mass spectrometry to a separation technique such as liquid chromatography or ion mobility spectrometry. This study focused on the complete characterization of a nonionic surfactant using both LC-MS and IM-MS and utilized for the first time interactive (i.e., adsorption-mode) UPLC-MS for the analysis of an amphiphilic polymer. The LCMS and MS/MS data indicated that the surfactant investigated contains no other fatty acid besides stearic acid; if a natural fat, such as palm oil, were used in the synthesis, admixtures of other fatty esters would have been detected. This information would be difficult to obtain by MS and MS/MS via direct infusion alone given the complexity of the analyte and the number of isobaric components present in the sample. Meanwhile, the IMMS analysis identified components in PEOn-glucam sesquistearate that were not detected by UPLC-MS and direct infusion. Specifically, the presence of small amounts of unesterified PEOn-glucam and of two disaccharide species, PEOn-diglucam stearate and PEOn-diglucam distearate, in the surfactant sample was conclusively confirmed. Conversely, IMMS analysis did not detect the most hydrophobic component, PEOn-glucam tristearate, which was seen by UPLC-MS. Thus, each approach offers complementary information, through its ability to separate minor constituents, by polarity (in LC-MS) or shape/size (in IM-MS), from more readily ionizable and/or major constituents, so that they can be characterized unequivocally without interferences. An advantage of IM-MS over LC-MS is the much faster analysis time and its minimal solvent requirements. Nevertheless, the most complete characterization of major and minor components of a surfactant, which would enable the differentiation of very similar samples of distinct origin, necessitates the use of both methods. LC separation of the polar compounds in amphiphilic systems, such as the unesterified PEOn and PEOn-glucam or the disaccharide components in the nonionic surfactant studied, could be improved by using hydrophilic interaction liquid chromatography (HILIC), which employs polar stationary phases (as in normal-phase LC) and water-based eluents (as in reverse-phase LC), and where analytes are eluted in order of rising polarity by increasing the water content of the mobile phase.14,57 Further sensitivity and dynamic range enhancement can be achieved by combining LC and IM separation in 3-D LC-IM-MS experiments. These methodologies will be explored in future studies.

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: (330) 972-7699. Fax: (330) 972-6085. Present Addresses †

B.C.K.: The Valspar Corporation, Global Technical Center, 2001 Tracy Street, Pittsburgh, PA 15233, U.S.A. ‡ S.F.H.: University of South Florida, College of Medicine, 12901 Bruce B Downs Blvd., Tampa, FL 33612, U.S.A. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The National Science Foundation is gratefully acknowledged for generous financial support (grant CHE-1308307). S.H. was supported by an REU internship funded by the National Science Foundation (grant DMR-1004747). We thank The Lubrizol Corporation (Wickliffe, OH) for the PEOn-glucam sesquistearate sample and Selim Gerişlioğlu and Kevin J. Endres for experimental assistance and helpful discussions.

■

REFERENCES

(1) Plat, T.; Linhardt, R. J. J. Surfactants Deterg. 2001, 4, 415−421. (2) Gu, L.; Faig, A.; Abdelhamid, D.; Uhrich, K. Acc. Chem. Res. 2014, 47, 2867−2877. (3) Kerwin, B. A. J. Pharm. Sci. 2008, 97, 2924−2935. (4) Jiao, J. Adv. Drug Delivery Rev. 2008, 60, 1663−1673. (5) Beneito-Cambra, M.; Herrero-Martinez, J. M.; Ramis-Ramos, G. Anal. Methods 2013, 5, 341−354. (6) Heinig, K.; Vogt, C.; Werner, G. J. Chromatogr. A 1996, 745, 281−292. (7) Kibbey, T. C. G.; Yavaraski, T. P.; Hayes, K. F. J. Chromatogr. A 1996, 752, 155−165. (8) Boyd-Boland, A. A.; Pawliszyn, J. B. Anal. Chem. 1996, 68, 1521− 1529. (9) Mico-Tormos, A.; Simo-Alfonso, E. F.; Ramis-Ramos, G. J. Sep. Sci. 2010, 33, 1398−1404. (10) Park, H. S.; Rhee, C. K. J. Chromatogr. A 2004, 1046, 289−291. (11) Sparreboom, A.; Zhao, M.; Brahmer, J. R.; Verweij, J.; Baker, S. D. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2002, 773, 183− 190. (12) Borisov, O. V.; Ji, J. A.; Wang, Y. J.; Vega, F.; Ling, V. T. Anal. Chem. 2011, 83, 3934−3942. (13) Hewitt, D.; Alvarez, M.; Robinson, K.; Ji, J.; Wang, Y. J.; Kao, Y. H.; Zhang, T. J. Chromatogr. A 2011, 1218, 2138−2145. (14) Abrar, S.; Trathnigg, B. Anal. Bioanal. Chem. 2011, 400, 2119− 2130. (15) Jandera, P.; Holčapek, M.; Theodoridis, G. J. Chromatogr. A 1998, 813, 299−311. (16) Levine, L. H.; Garland, J. L.; Johnson, J. V. Anal. Chem. 2002, 74, 2064−2071. (17) Desmazières, B.; Buchmann, W.; Terrier, P.; Tortajada, J. Anal. Chem. 2008, 80, 783−792. (18) Nasioudis, A.; van Velde, J. W.; Heeren, R. M. A.; van den Brink, O. F. J. Chromatogr. A 2011, 1218, 7166−7172. (19) Deng, G.; Chow, D.; Sanyal, G. Anal. Biochem. 2001, 289, 124− 129. (20) Zgoła-Grześkowiak, A.; Grześkowiak, T.; Frańska, M.; Rząsa, A.; Łukaszewski, Z. Biodegradation 2008, 19, 635−642. (21) Van Leeuwen, S. M.; Tan, B.; Grijpma, D. W.; Feijen, J.; Karst, U. Rapid Commun. Mass Spectrom. 2007, 21, 2629−2637. (22) Song, J.; Šišková, A.; Simons, M. G.; Kowalski, W. J.; Kowalczuk, M. M.; van den Brink, O. F. J. Am. Soc. Mass Spectrom. 2011, 22, 641− 648. (23) McEwen, C. N.; McKay, R. G.; Larsen, B. S. Anal. Chem. 2005, 77, 7826−7831.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b03400. Compositional tables; mass spectra extracted from LC fractions and IM-MS bands; MS/MS spectra of select oligomers; fragmentation mechanisms (PDF) F

DOI: 10.1021/acs.analchem.5b03400 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry (24) Snelling, J. R.; Scarff, C. A.; Scrivens, J. H. Anal. Chem. 2012, 84, 6521−6529. (25) Solak Erdem, N.; Alawani, N.; Wesdemiotis, C. Anal. Chim. Acta 2014, 808, 83−93. (26) Pasch, H.; Trathnigg. B. HPLC of Polymers; Springer: Berlin, Germany, 1998. (27) Murgasova, R.; Hercules, D. M. Anal. Bioanal. Chem. 2002, 373, 481−489. (28) Aaserud, D. J.; Prokai, L.; Simonsick, W. J., Jr. Anal. Chem. 1999, 71, 4793−4799. (29) Zhu, J. J. Surfactants Deterg. 2004, 7, 421−423. (30) Levine, L. H.; Garland, J. L.; Johnson, J. V. J. Chromatogr. A 2005, 1062, 217−225. (31) Scionti, V.; Katzenmeyer, B. C.; Solak Erdem, N.; Li, X.; Wesdemiotis, C. Eur. Mass Spectrom. 2012, 18, 113−137. (32) Swartz, M. E. J. Liq. Chromatogr. Relat. Technol. 2005, 28, 1253− 1263. (33) Falkenhagen, J.; Weidner, S. Anal. Chem. 2009, 81, 282−287. (34) Bowers, M. T.; Kemper, P. R.; von Helden, G.; van Koppen, P. A. Science 1993, 260, 1446−1451. (35) Clemmer, D. E.; Jarrold, M. F. J. Mass Spectrom. 1997, 32, 577− 592. (36) Trimpin, S.; Plasencia, M.; Isailovic, D.; Clemmer, D. E. Anal. Chem. 2007, 79, 7965−7974. (37) Trimpin, S.; Clemmer, D. E. Anal. Chem. 2008, 80, 9073−9083. (38) Kanu, A. B.; Dwivedi, P.; Tam, M.; Matz, L.; Hill, H. H. J. Mass Spectrom. 2008, 43, 1−22. (39) Hilton, G. R.; Jackson, A. T.; Thalassinos, K.; Scrivens, J. H. Anal. Chem. 2008, 80, 9720−9725. (40) Chan, Y.-T.; Li, X.; Soler, M.; Wang, J.-L.; Wesdemiotis, C.; Newkome, G. R. J. Am. Chem. Soc. 2009, 131, 16395−16397. (41) Brocker, E. R.; Anderson, S. E.; Northrop, B. H.; Stang, P. J.; Bowers, M. T. J. Am. Chem. Soc. 2010, 132, 13486−13494. (42) Chan, Y.-T.; Li, X.; Yu, J.; Carri, G. A.; Moorefield, C. N.; Newkome, G. R.; Wesdemiotis, C. J. Am. Chem. Soc. 2011, 133, 11967−11976. (43) Li, X.; Guo, L.; Casiano-Maldonado, M.; Zhang, D.; Wesdemiotis, C. Macromolecules 2011, 44, 4555−4564. (44) Li, X.; Chan, Y.-T.; Casiano-Maldonado, M.; Carri, G. A.; Newkome, G. R.; Wesdemiotis, C. Anal. Chem. 2011, 83, 6667−6674. (45) Scarff, C. A.; Snelling, J. R.; Knust, M. M.; Wilkins, C. L.; Scrivens, J. H. J. Am. Chem. Soc. 2012, 134, 9193−9198. (46) Yol, A. M.; Wesdemiotis, C. React. Funct. Polym. 2014, 80, 95− 108. (47) Guo, K.; Guo, Z.; Ludlow, J. M., III; Xie, T.; Liao, S.; Newkome, G. R.; Wesdemiotis, C. Macromol. Rapid Commun. 2015, 36, 1539− 1552. (48) De Hoffmann, E.; Stroobant, V. Mass Spectrometry: Principles and Applications, 3rd ed.; John Wiley & Sons Ltd: Chichester, West Sussex, England, 2007. (49) Puglisi, C.; Samperi, F.; Carroccio, S.; Montaudo, G. Rapid Commun. Mass Spectrom. 1999, 13, 2268−2277. (50) Chaicharoen, K.; Polce, M. J.; Singh, A.; Pugh, C.; Wesdemiotis, C. Anal. Bioanal. Chem. 2008, 392, 595−607. (51) Wesdemiotis, C.; Solak, N.; Polce, M. J.; Dabney, D. E.; Chaicharoen, K.; Katzenmeyer, B. C. Mass Spectrom. Rev. 2011, 30, 523−559. (52) Katzenmeyer, B. C. Interfacing liquid chromatography or ion mobility separation with multidimensional mass spectrometry for the structural characterization of polymeric materials. Ph.D. Dissertation, The University of Akron, 2013. (53) Pringle, S. D.; Giles, K.; Wildgoose, J. L.; Williams, J. P.; Slade, S. E.; Thalassinos, K.; Bateman, R. H.; Bowers, M. T.; Scrivens, J. H. Int. J. Mass Spectrom. 2007, 261, 1−12. (54) Jurneczko, E.; Barran, P. E. Analyst 2011, 136, 20−28. (55) Yu, T.; Traina, J. A.; Pungor, E., Jr.; McCaman, M. Anal. Biochem. 2006, 359, 54−62. (56) Yoo, C.; Suckau, D.; Sauerland, V.; Ronk, M.; Ma, M. J. Am. Soc. Mass Spectrom. 2009, 20, 326−333.

(57) Liu, X.; Pohl, C. J. Chromatogr. A 2008, 1191, 83−89.

G

DOI: 10.1021/acs.analchem.5b03400 Anal. Chem. XXXX, XXX, XXX−XXX