Anal. Chem. 1998, 70, 3220-3226
Identification, Composition, and Asymmetric Formation Mechanism of Glycidyl Methacrylate/ Butyl Methacrylate Copolymers up to 7000 Da from Electrospray Ionization Ultrahigh-Resolution Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Stone D.-H. Shi,†,‡ Christopher L. Hendrickson,‡ Alan G. Marshall,*,†,‡ William J. Simonsick, Jr.,*,§ and David J. Aaserud§
Center for Interdisciplinary Magnetic Resonance, National High Magnetic Field Laboratory, and Department of Chemistry, Florida State University, Tallahassee, Florida 32310, and DuPont Automotive Finishes, Marshall R&D Laboratory, Philadelphia, Pennsylvania 19146
Glycidyl methacrylate (GMA) and butyl methacrylate (BMA) have the same nominal mass (142 Da) but differ in exact mass by 0.036 Da (CH4 vs O). Therefore, copolymers formed from the two isobaric monomers exhibit a characteristic isobaric distribution due to different monomer compositions. Here, we show that electrospray ionization FT-ICR mass spectrometry at 9.4 T resolves the isobaric components of copolymers as large as 7000 Da with a resolving power (m/∆m50%) of ∼500 000 in a gel permeation chromatography fractionated polymer sample. That resolution provides for complete and unequivocal component analysis of such copolymers of the size used for high solid content automobile coatings. All five possible copolymer products predicted by the polymerization mechanism are resolved and identified in the mass spectrum. Two of those polymer series (each with saturated end group) were previously unresolved by mass spectrometry because they differ in mass from the two other unsaturated products by only 0.0089 Da. Finally, analysis of the asymmetrical isobaric distribution for the copolymer n-mers, (GMA)m(BMA)n-m, 0e m e n, in which species with adjacent values of m differ from each other in mass by 36 mDa (i.e., the mass difference, CH4 vs O, between GMA and BMA) proves that GMA is less reactive than BMA in the polymerization process. Mass spectrometry is widely used to characterize synthetic polymers, because a mass spectrum contains the complete distribution of polymer molecules (chain length, monomer composition, and end groups),1-8 not just the number average (as from colligative properties) or weight average (as from sedimentation †
Department of Chemistry, Florida State University. National High Magnetic Field Laboratory. § DuPont Automotive Finishes. (1) Bahr, U.; Deppe, A.; Karas, M.; Hillenkamp, F. Anal. Chem. 1992, 64, 28662869. (2) Smith, P. B.; Pasztor, A. J., Jr.; McKelvy, M. L.; Meunier, D. M.; Froelicher, S. W.; Wang, F. C.-Y. Anal. Chem. 1997, 69, 95R-121R. ‡
3220 Analytical Chemistry, Vol. 70, No. 15, August 1, 1998
or light scattering).9 Matrix-assisted laser desorption/ionization (MALDI) is often used to characterize molecular weight distributions.3,5-7,10-17 MALDI-TOF (time-of-flight) analysis of the molecular weight distribution has been shown to agree with the result from traditional methods such as gel permeation chromatography (GPC) for samples of low polydispersity.18-20 The main advantage of MALDI is that it tends to produce singly charged ions, thereby yielding relatively few ionic species and a correspondingly simple mass spectrum. However, the mass-to-charge ratio is necessarily high, and it is, therefore, very difficult to resolve isotopic (let alone isobaric) species for polymers of more than a few thousand daltons.5,6,21,22 (3) Brown, R. S.; Weil, D. A.; Wilkins, C. L. Macromolecules 1986, 19, 12551260. (4) Nuwaysir, L. M.; Wilkins, C. L.; Simonsick, W. J. J. Am. Soc. Mass Spectrom. 1990, 1, 66-71. (5) Dey, M.; Castoro, J. A.; Wilkins, C. L. Anal. Chem. 1995, 67, 1575-1579. (6) Pastor, S. J.; Wilkins, C. L. J. Am. Soc. Mass Spectrom. 1997, 8, 225-233. (7) O’Connor, P. B.; Duursma, M. C.; Vanrooij, G. J.; Heeren, R. M. A.; Boon, J. J. Anal. Chem. 1997, 69, 2751-2755. (8) Jackson, C. A.; Simonsick, W. J., Jr. Curr. Opin. Solid State Mater. Sci. 1997, 2, 661-667. (9) Marshall, A. G. Biophysical Chemistry: Principles, Techniques, and Applications; Wiley: New York, 1978. (10) Karas, M.; Bahr, U.; Giessman, U. Mass Spectrom. Rev. 1991, 10, 335357. (11) Ko ¨ster, C.; Kahr, M. S.; Castoro, J. A.; Wilkins, C. L. Mass Spectrom. Rev. 1992, 11, 495-512. (12) de Koster, C.; Duursma, M. C.; van Rooij, G. J.; Heeren, R. M. A.; Boon, J. J. Rapid Commun. Mass Spectrom. 1995, 9, 957-962. (13) Vanderhage, E. R. E.; Duursma, M. C.; Heeren, R. M. A.; Boon, J. J.; Nielen, M. W. F.; Weber, A. J. M.; de Koster, C. G.; Devries, N. K. Macromolecules 1997, 30, 4302-4309. (14) Heeren, R. M. A.; Boon, J. J. Int. J. Mass Spectrom. Ion Processes 1996, 158, 391-403. (15) van Ronij, G. J.; Duursma, M. C.; Heeren, R. M. A.; Boon, J. J.; de Koster, C. G. J. Am. Soc. Mass Spectrom. 1996, 7, 449-457. (16) McEwen, C. N.; Jackson, C.; Larsen, B. S. Int. J. Mass Spectrom. Ion Processes 1997, 160, 387-394. (17) Vitalini, D.; Mineo, P.; Scamporrino, E. Macromolecules 1997, 30, 52855289. (18) Lloyd, P. M.; Suddaby, K. G.; Varney, J. E.; Scrivener, E.; Derrick, P. J.; Haddleton, D. M. Eur. Mass Spectrom. 1995, 1, 293-300. (19) Guttman, C. M. Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 1996, 37, 837-838. (20) Jackson, C.; Larsen, B.; McEwen, C. Anal. Chem. 1996, 68, 1303-1308. S0003-2700(98)00316-3 CCC: $15.00
© 1998 American Chemical Society Published on Web 06/18/1998
Alternatively, electrospray ionization (ESI)23-25 can yield multiply charged ions at relatively low mass-to-charge ratio (m/z e 2000), for which potentially ultrahigh mass resolving power is available with Fourier transform ion cyclotron resonance (FT-ICR) mass analysis.26-29 Possible fragmentation due to metastable dissociation of internally hot ions observed in MALDI experiments30 is also less likely to occur when using ESI.31 However, the presence of multiple charge states complicates the mass spectrum, and the resulting dynamic range limitation can mitigate the realization of high mass resolution by FT-ICR MS, as demonstrated by a recent analysis of poly(ethylene glycol), in which more than 5000 species were present in a sample of ∼20 kDa average molecular mass.32 Automated charge state deconvolution33-36 can help simplify such complex spectra. Coupling (either on-line or off-line) of mass spectrometry with separation methods makes possible analysis of very complex polymer samples. GPC fractionation narrows the molecular weight distribution in each fraction so that accurate mass measurement can be performed by MALDI MS.37-40 Coupling GPC with ESI MS41-43 simplifies the spectrum of each fraction and, as shown below, also facilitates analysis of the complete mass distribution of the polymer sample. Polymers with molecular mass up to 5 MDa have been detected (but not resolved) by ESI mass spectrometry.44,45 In addition to the molecular weight distribution,20,46 potentially much (21) Easterling, M. L.; Pitsenberger, C. C.; Kulkarni, S. S.; Taylor, P. K.; Amster, I. J. Int. J. Mass Spectrom. Ion Processes 1996, 158, 97-113. (22) Pitsenberger, C. C.; Easterling, M. L.; Amster, J. I. Anal. Chem. 1996, 68, 3732-3739. (23) Whitehouse, C. M.; Dreyer, R. N.; Yamashita, M.; Fenn, J. B. Anal. Chem. 1985, 57, 675-679. (24) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 246, 64-71. (25) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F. Mass Spectrom. Rev. 1990, 9, 37-70. (26) McLafferty, F. W. Acc. Chem. Res. 1994, 27, 379-386. (27) Senko, M. W.; Hendrickson, C. L.; Pasa-Tolic, L.; Marto, J. A.; White, F. M.; Guan, S.; Marshall, A. G. Rapid Commun. Mass Spectrom. 1996, 10, 1824-1828. (28) Kelleher, N. L.; Senko, M. W.; Siegel, M. M.; McLafferty, F. W. J. Am. Soc. Mass Spectrom. 1997, 8, 380-383. (29) Marshall, A. G.; Hendrickson, C. L.; Jackson, G. S. Mass Spectrom. Rev. In press. (30) Feast, W. J.; Hamilton, L. M.; Rannard, S. Polym. Bull. 1997, 39, 347-352. (31) McEwen, C. N.; Simonsick, W. J., Jr.; Larsen, B. S.; Ute, K.; Hatada, K. J. Am. Soc. Mass Spectrom. 1995, 6, 906-911. (32) O’Connor, P. B.; McLafferty, F. W. J. Am. Chem. Soc. 1995, 117, 1282612831. (33) Zhang, Z.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 1998, 9, 225-233. (34) Mann, M.; Meng, C. K.; Fenn, J. B. Anal. Chem. 1989, 61, 1702-1708. (35) Reinhold, B. B.; Reinhold, V. N. J. Am. Soc. Mass Spectrom. 1992, 3, 207215. (36) Senko, M. W.; Beu, S. C.; McLafferty, F. W. J. Am. Soc. Mass Spectrom. 1995, 6, 52-56. (37) Montaudo, G.; Garozzo, D.; Montaudo, M. S.; Puglisi, C.; Samperi, F. Macromolecules 1995, 28, 7983-7989. (38) Montaudo, G.; Montaudo, M. S.; Puglisi, C.; Samperi, F. Rapid Commun. Mass Spectrom. 1995, 9, 1158-1163. (39) Montaudo, G.; Montaudo, M. S.; Puglisi, C.; Samperi, F. Int. J. Polym. Anal. Charact. 1997, 3, 177-192. (40) Fei, X.; Murray, K. K. Anal. Chem. 1996, 68, 3555-3560. (41) Prokai, L.; Simonsick, W. J., Jr. Rapid Commun. Mass Spectrom. 1993, 7, 853-6. (42) Simonsick, W. J., Jr.; Prokai, L. In Chromatographic Characterization of Polymers: Hyphenated and Multidimensional Techniques; Provder, T., Barth, H. G., Urban, M., Eds.; ACS Advances in Chemistry Series 247; American Chemical Society: Washington, DC, 1995; pp 41-56. (43) Nielen, M. W. F. Rapid Commun. Mass Spectrom. 1996, 10, 1652-1660. (44) Nohmi, T.; Fenn, J. B. J. Am. Chem. Soc. 1992, 114, 3241-3246.
more information can be extracted by ultrahigh-resolution mass analysis. For example, isobaric monomers such as glycidyl methacrylate (GMA) and butyl methacrylate (BMA) have the same “nominal” mass (142 Da for GMA and BMA) but differ in exact mass (in this case, by 0.036 Da, the mass difference between CH4 and O). Although one could attempt to distinguish the two monomers in a GMA/BMA copolymer by selective reaction of one monomer type to cause a mass shift,47 a simpler approach is to distinguish the isobaric peaks directly by ultrahigh-resolution mass spectrometry, to determine the relative abundances of n-mer individual isobaric components, (GMA)m(BMA)n-m, 0e m e n. Although isobaric peaks could not be resolved at all by MALDI FT-ICR MS at 3.0 T,47 ESI FT-ICR MS at 3.0 T could resolve isobaric species from small oligomers (up to the hexamer, ∼852 Da).48 Online coupling of GPC with FT-ICR mass analysis successfully analyzed this copolymer up to the 48-mer (MW < 7000) with isotopic resolution (i.e., to within 1 Da) but not isobaric resolution.49 Similar copolymers containing reactive monomers such as GMA have potential applications for high solid content automobile coatings,50 for which the upper molecular mass is kept low (under 10 kDa) to avoid high viscosity. Here, we show that ESI ultrahigh-resolution FT-ICR MS extends the determination of copolymer composition (and formation mechanism) to oligomers up to 7000 Da. Only FT-ICR MS29 can provide the mass resolving power needed for that purpose. For example, isotope resolution for a protein in excess of 100 kDa has been achieved by ESI FT-ICR MS at 9.4 T.28 High magnetic field, B, is essential, because mass resolving power is proportional to B,51 whereas the tendency for closely spaced resonances to coalescence varies inversely with B.52 We therefore employ a 9.4-T ESI FT-ICR instrument27,53 for the present copolymer analysis. As we shall show, a complete picture of the polymerization reaction mechanism is revealed by detection of all five predicted polymer product series, two of which were not previously observed due to insufficient mass resolving power. We also establish a lower reactivity for GMA relative to BMA in the polymerization process, based on relative abundances of isobaric copolymers. EXPERIMENTAL METHODS Sample Preparation. The 50/50 (w/w) glycidyl methacrylate (GMA)/n-butyl methacrylate (BMA) macromonomer was prepared in the following manner. In a 5-L glass flask, 0.21 g of Co(dimethylglyoxime-BF2)2 dissolved in 343.12 g of methyl ethyl (45) Bruce, J. E.; Cheng, X.; Bakhtiar, R.; Wu, Q.; Hofstadler, S. A.; Anderson, G. A.; Smith, R. D. J. Am. Chem. Soc. 1994, 116, 7839-7847. (46) Latourte, L.; Blais, J.-C.; Tabet, J.-C.; Cole, R. B. Anal. Chem. 1997, 69, 2742-2750. (47) Ross, C. W., III; Simonsick, W. J., Jr. Proceedings of the 44th ASMS Conference on Mass Spectrometry and Allied Topics; Portland, OR, May 12-16, 1996; p 1270. (48) Simonsick, W. J., Jr.; Aaserud, D. J.; Grady, M. C.; Prokai, L. Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 1997, 38, 483-484. (49) Aaserud, D. J.; Prokai, L.; Simonsick, W. J., Jr. Proceedings of the 45th ASMS Conference on Mass Spectrometry and Allied Topics; Palm Springs, CA, June 1-5, 1997; p 1283. (50) Weiss, K. D. Prog. Polym. Sci. 1997, 22, 203-245. (51) Marshall, A. G.; Guan, S. Rapid Commun. Mass Spectrom. 1996, 10, 18191823. (52) Mitchell, D. W.; Smith, R. D. Phys. Rev. E 1995, 52, 4366-4386. (53) Senko, M. W.; Hendrickson, C. L.; Emmett, M. R.; Shi, S. D.-H.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 1997, 8, 970-976.
Analytical Chemistry, Vol. 70, No. 15, August 1, 1998
3221
ketone (MEK) was heated to reflux (80 °C) under a nitrogen blanket. To that mixture, 1050 g of GMA and 1050 g of BMA were added over a 4 h period. Concurrent with the monomer feed, 42 g of the initiator, Vazo-52 (2, 2′-azobis[2,4-dimethylpentanenitrile], DuPont Specialty Chemicals, Wilmington, DE) dissolved in 514.7 g of MEK was added over 5 h. The mixture was held at reflux for 30 min after the feeds and subsequently allowed to cool to room temperature. A more detailed procedure for the macromonomer preparation and the Co(dimethylglyoxime-BF2)2 synthesis is available.54 The polymer sample can be analyzed directly by FT-ICR MS by use of the methods described below. For better detection of the high-molecular-weight components in the sample, fractionation by gel permeation chromatography (GPC) was carried out with a four-column set of two 1000-, one 500-, and one 100-Å, 30-cm × 7.8-mm-i.d. Ultrastyragel columns (Waters, Milford, MA). The tetrahydrofuran (THF) mobile phase was delivered by a HewlettPackard (Palo Alto, CA) model 1050 gradient liquid chromatograph at 1.0 mL/min. The analyte, dissolved in the mobile phase (∼1% w/v), was injected via a 100-µL loop. Data from the HewlettPackard model 1037E refractive index (RI) detector were collected with a Waters Millennium system. The chromatogram, together with the fraction cutoff, is shown in Figure 6 (see below). The fractions collected from GPC were dried and then redissolved for FT-ICR MS analysis. The solution for ESI FT-ICR was made by first adding THF to the sample (GMA/BMA copolymer or its dried GPC fractions) and then adding the same volume of methanol. Sufficient 10 mM NaI solution in methanol was added to bring the salt concentration to 0.1 mM. The concentration of the copolymer in the solution was kept between 1 and 10 mg/mL. The concentrations of the GPC fraction samples are estimated as 0.5-5 mg/mL. Mass Analysis. FT-ICR experiments were performed with a home-built 9.4-T FT-ICR mass spectrometer27,53 controlled by an Odyssey data station (Finnigan FTMS, Madison, WI). The ions generated from a microelectrospray source55 were accumulated external to the magnet53 in a 67-cm-long rf-only octopole for 5-10 s and then transferred through a 200-cm-long rf-only octopole into a 10-cm-i.d. open cylindrical ICR cell (10 cm diameter, 30 cm long). The pressure reading (noncalibrated) on the ion gauge was kept below 7 × 10-9 Torr in the Penning trap. Ions were detected following broadband dipolar radial excitation (frequency sweep rate, 300 Hz/µs at peak-to-peak amplitude of 124.5 V). Broadband time domain data (512 Kword) were acquired at a Nyquist frequency of 222 222 Hz (corresponding to a lower mass-to-charge ratio limit of 650). Ultrahigh-resolution FT-ICR mass spectra were obtained by isolating the ions of interest by stored waveform inverse Fourier transform (SWIFT56,57) radial dipolar mass-selective ejection, followed by conventional frequency sweep (chirp) radial dipolar excitation (300 Hz/µs at 124.5 Vp-p) with heterodyne detection. Heterodyne detection provides for the required 5-15-s time domain signal acquisition period without generating an inconve(54) Janowicz, A. H.; Melby, L. R. U.S. Patent 4,680,352, 14 Jan 1987. (55) Emmett, M. R.; White, F. M.; Hendrickson, C. L.; Shi, S. D.-H.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 1998, 9, 333-340. (56) Marshall, A. G.; Wang, T.-C. L.; Ricca, T. L. J. Am. Chem. Soc. 1985, 107, 7893-7897. (57) Guan, S.; Marshall, A. G. Int. J. Mass Spectrom. Ion Processes 1996, 157/ 158, 5-37.
3222 Analytical Chemistry, Vol. 70, No. 15, August 1, 1998
niently large data set. Reference frequency and detection bandwidth of the heterodyne detection varied with the required detection window. For example, GPC fraction 4 was detected at a reference frequency of 68 702 Hz and a Nyquist frequency of 9259 Hz (corresponding to a mass-to-charge ratio window of 21002427), to yield 256 Kword time domain data. A low trapping potential, typically 1.0 V on each end cap electrode, minimized frequency shift and drift58,59 during the long detection event (typically 10 s) required for ultrahigh mass resolving power. The data were processed by use of in-house FT-ICR data analysis software.60 One zero-fill and a Hanning apodization were applied to the data before magnitude mode Fourier transform. A total of 70-100 time domain data sets from each GPC fraction were coadded to enhance the signal-to-noise ratio. For accurate mass measurements, an internal mass calibrant, poly(ethylene glycol) bis(carboxymethyl) ether (number average molecular weight, Mn ) 600, Aldrich, Milwaukee, WI), was added to the sample solution. The concentration of the calibrant was adjusted to match the calibrant and analyte ion abundances. The analyte ions, together with the calibrant ions of the two closest m/z values, were isolated by SWIFT mass-selective ejection. The calibration is thus limited to calibrant ions with m/z values very close to those of the analyte, thereby improving mass measurement accuracy for two reasons. First, the calibrant and the analyte ions experience exactly the same electric and magnetic fields. Second, because the calibrant is close in m/z value to the analyte, deviation due to imperfect fit of the calibration equation61 to the calibration curve is minimized. The molecular weight of each calibrant ion was calculated from its elemental formula,62 taking care to subtract the mass of one electron (0.000 548 57 Da) from each neutral molecule to generate an accurate mass-to-charge ratio for its corresponding cation. (The electron mass is, of course, added back to the final analyte ion mass to yield the reported mass of each neutral molecule.) Each measurement was repeated five times, to give a reproducibility of
