Polymeric materials, ubiquitous in modern society, are constituted by individual macromolecular chains (sometimes crosslinked), with the material properties dictated by the nature and architecture of the individual constituent chains.
Important parameters governing material properties include molecular weight (Mn) and architecture (linear, branched, star, cyclic, etc), monomer composition, degree of and sites of functionalization, and the presence of impurities or additives. In recent decades, the introduction of controlled/living free radical polymerization techniques has revolutionized the engineering of individual macromolecular chains by imparting high levels of mechanistic control over the polymerization process.1–14 In addition, the ability of polymer scientists to design and precisely synthesize complex macromolecules has been enhanced by the recent resurgence of “click” chemical transformations.15–17 Often very small percentage changes made to polymer chains results in significant property changes; hence, end-groups become crucial in determining chain degradation and/or ensuring reactivation in living polymerization reactions. As end-groups of macromolecules constitute a very small % of the overall chain, this imposes constraints on accurate analyses via conventional spectroscopic approaches (e.g., NMR, where complex radioactive labeling or end-group transformation chemistry has often been used to enhance signals). Refinements in polymer synthesis that allow for greater control over the chains, can only have a proven benefit if analytical tools are codeveloped to allow for careful verification of the products. Prior to the development of MS with soft ionization, for example, MALDI or ESI-MS, mass spectrometry had limited appeal as a technique for providing insights into polymerization reaction mechanisms as analysis without chain degradation (in the spectrometer) was very difficult to achieve. MS studies were mainly limited to the study of polymer degradation by (for example) coupling to pyrolysis gas chromatography with mass spectrometric analysis.
Mass Spectrometry has the potential to provide unprecedented access to mechanistic polymerization insights, as it provides separation at the single chain scale with extremely high accuracy, provided chain stability on ionization, and detection can be assured, advances that have now become reality using soft-ionization.
Soft ionization MS techniques, electrospray ionization (ESI) and soft laser desorption, often known as matrix assisted laser desorption ionization (MALDI), have had a significant impact on polymer analytical chemistry (as well as on protein and peptide analyses where the techniques are still more prevalently used). The development of ESI and MALDI resulted in the Nobel Prize in Chemistry being awarded to John B. Fenn and Koichi Tanaka in 2002.18–23 ESI and MALDI will be the focus of this review as the application of these techniques has revolutionized polymer analytical chemistry facilitating major advances in mechanistic understanding of both polymerization and depolymerisation (or degradation) processes and mechanisms. A brief search in SciFinder scholar with keywords “polymer,” “mass spectrometry,” and “degradation,” including English journal research articles published between 2005–2012 gave over 1300 references. Therefore, we have judiciously included only references that have special relevance to polymerization. In particular, the main focus of this review will be on the MS analyses of degraded and chemically modified polymer chains, as emphasis on using MS to study polymerization mechanisms has been published in a prior review.24–26
Soft Ionization Technique
Since the introduction of soft ionization MS techniques, such as ESI and MALDI, have become routine analytical tools for polymer analyses complementing more established analytical approaches such as NMR and IR spectroscopy and Size exclusion chromatography, all techniques incapable of single chain resolution and/or isolation.24, 25, 27, 28
Soft ionization involves the transformation of a solid or dissolved liquid sample into the gas phase, without inducing any changes to the molecular structure of the sample. In contrast early mass spectrometry used “hard ionization” methods including, electron ionization (EI), chemical ionization (CI), fast atom bombardment (FAB), etc, which destroy the chemical structures into massive fragments prior to detection by mass spectrometry. In MALDI, ionization is achieved by firing a laser into a matrix comixed with sample analyte. The matrix is designed to absorb the vast majority of the laser energy inducing ionization of the matrix, softening the energy transfer to analyte, providing a gentle ionization process as shown by Figure 1. The purpose of the matrix is twofold, it helps ionize the sample analyte while also protecting it from the direct energy of the laser thereby minimizing or eliminating fragmentation processes. MALDI is often coupled with a time-of-flight (TOF) mass spectrometer. The TOF facilitates the measurement of chains produced from individual (or collated) laser “shots.”
MALDI–MS techniques can be exploited for the analyses of a large mass range, especially if the instrumentation is coupled with an “ion mirror” to deflect the ions with an electric field; effectively doubling the potential flight path providing increased resolution.
The ESI-MS, ionization process is based on the exploitation of repulsive electric forces to disperse sample solutions into a fine aerosol form. The aerosol then undergoes a solvent evaporation stage during which it becomes unstable as it reaches the Rayleigh limit. The droplets will then burst and form charged jets in a process called coulomb fission. Finally, the sample forms gas phase ions on entering the mass spectrometer via one of two possible mechanisms (Ion Evaporation Model and Charge Residue Model) (Fig. 2). All types of mass analyzers have been used with ESI including TOF, quadrupole, ion trap, orbitrap, and Fourier transform ion cyclotron resonance (FT-ICR). Each mass spectrometer has different resolving power, mass accuracy, and dynamic range as outlined in Table 1. Of all mass spectrometers available, the most powerful are FTICR and Orbitrap, often coupled with a linear ion trap. It is important to note that the orbitrap, invented by Alexander Makarov earlier this decade may become critically important in future applications as it is capable of high resolution (mass accuracy within 2 ppm), together with maximum resolving power of 100,000 (full width at half maximum) and operational simplicity.30–32 Although this operational capability is still much lower than FTICR (resolving power >1,000,000), it is sufficient for most applications in polymer science. A limitation to FTICR is the requirement for a superconducting magnet imposing the costly need for large specifically designed spaces. In contrast, the orbitrap mass spectrometer is much smaller, simpler, and cheaper.31, 33
Table 1. Typical Performance of Mass Analyzer Employed for MALDI and ESI-MS
ESI and MALDI, both generate quasi-molecular ions carrying a unit charge (either + or −). The most common counter ions observed/employed are protons [M + H]+ and sodium [M + Na]+ although there are also other types of adducts that are sometimes chosen for different polymer families (vide infra). Besides ESI and MALDI, it is also important to note that there are other ionization techniques that are currently under development exploiting the basic principles of ESI and MALDI, for example, TOF secondary ion mass spectrometry (TOF-SIMS), desorption electrospray ionization (DESI), electrosonic spray ionization (ESSI), and atmospheric pressure chemical ionization (APCI) techniques known to be better for less polar analytes. The use of APCI for the analyses of synthetic polymers has received scant attention.
Correct sample preparation is essential for effective and reliable mass spectrometric analyses, (more so for MALDI-MS than for ESI). In MALDI-MS, the choice and optimization of the matrix is critical. Important characteristics of MALDI matrices include matching solubility between sample and solvent, good miscibility with the analyte polymer, low vapor pressure, good vacuum stability, and high light absorption at the laser wavelength. An important characteristic of successful MALDI is the need to incorporate analyte molecules into the matrix crystalline structure to facilitate analyte transport into the gas phase following laser irradiation (hence, matrix assisted).34–36 Choosing the right matrix for a polymer can be an empirical exercise as the role of matrix is still not fully understood.37 As a result, the optimization procedure can involve a large number of trial and error experiments. A simple method to decide on the suitability of a matrix has been suggested by Hoteling et al.,38, 39 the matrix/polymer combination that has closest match of desorption time, results in the best MALDI spectrum. Alternatively, Hanton and Owens40, 41 have proposed that the best solid solution achieved is by the matching polarity of the matrix and analyte using matrix-enhanced secondary ion mass spectrometry. Ideally, a homogeneous analyte/matrix matrix mixture after solvent evaporation is achieved, as shown in Figure 3 where (a) a uniform (homogeneous) sample is shown and (b) Segregation effects (in-homogeneities) are observed following a small addition of THF. Thus solvent choice is critical.38, 39 MALDI can be divided into “solvent based” and “solvent free” preparation methods. In most cases, solvent based systems are used where the mixture of matrix, analyte, and salts is “spotted” onto MALDI targets and then dried. However, this method is not appropriate for some classes of polymer encouraging the development of a “solvent free” methodby Trimpin et al.42 This has been successfully applied to theMS analyses of poly(etherimide),43, 44 aromatic poly(amides),45 and poly(9,9-diphenylfluorene).46
A useful starting point for accessing information on published experimental details for MALDI-MS analyses is the NIST47 website and the summary by Nielsen48 published in 1999. In this review, we provide an extensive updated polymer/matrix list as there has been a significant increase in the amount of published experimental detail in recent years (both MALDI-MS and ESI-MS). An updated list of polymers with their associated sample preparation recipes is shown in Table 4. Some common matrices suitable for the analyses of synthetic polymers are outlined in Table 2. It is crucially important to employ an optimal matrix for each different polymer to obtain the best results. Matrix choice and preparation has been shown to affect greatly the quality of mass spectra obtained in terms of signal intensity and measurability, for exactly the same analyte. This fact is demonstrated in Figure 4.49 The choice of matrix is also dictated by the ion mode selection of the MALDI-MS (positive or negative ion modes), as shown by Wyatt et al.50 where DHB and THAP were both suitable for use in the positive mode, but THAP clearly proved a better choice in negative mode for PMAA containing a pentaerythritol tetra(3-mercaptopropionate) end group. Schubert and coworkers used inkjet printing in combinatorial studies to choose the right matrix recipe for MALDI-MS.51, 52 In some cases, matrix addition can also cause chemical modification of the analyte; such as modification of NHC-polyester spirocycle to free cyclic poly(N-Bu-glycine) due to reaction with dithranol. Importantly, this does not occur in ESI-MS.53 The fact that one matrix does not fit all polymers means that new matrices must be developed to improve the measurability of certain polymers. For instance Ameduri et al.54 reported that poly(vinylidene fluoride) could not be ionized with common matrices and required a specific matrix [2,3,4,5,6-pentafluorocinnamic acid (PFCA)] for poly(fluoro-olefines), including poly(vinylidene fluoride) and copoly((vinylidene fluoride)-co-(perfluorovinylether)).55 While Berthod et al.49 also developed an ionic liquid matrix to facilitate the ionization of various polymers as can be seen in Figure 4. High quality mass spectra are not only important to assist in assigning molecular formula, but also for molecular weight and polydispersity calculations as can be seen in Table 3.
Table 2. Common MALDI Matrices for MS Studies on Synthetic Polymers
Table 3. Average Molecular Weight (in Number and Weight) of Polymers Obtained by MALDI-MS Using Different Matrices and a Comparison with Manufacturer Defined Values (GPC standards)49
Table 4. Range of Polymers Analyzed via MALDI or ESI Mass Spectrometry with the Appropriate Matrix Combination for MALDI and Appropriate Counter Ion Salt for ESI Taken from Previously Published Studies
When preparing MALDI samples, often a cationizing agent is used.23 Common cationizing agents are lithium, sodium, potassium, and silver salts.24, 26, 28, 48, 56–61 In some cases, the addition of metal salts can cause fragmentation of certain weaker bonds in polymer structures (e.g., end groups). Halohydrination of epoxy resin has been observed when some halogenated sodium salts (NaI and NaBr) have been used as cationizing agents. This halohydrination is not seen with NaCl because chlorine is less nucleophilic than bromine or iodine.62 For polymers that are not easily ionizable, it may be necessary to chemically derivatized them prior to measurement.63–65 Instrument parameter settings can also play an important role in successful mass spectrometry measurement with parameters such as laser pulse rate, mass range, accelerating voltage, reflectron/linear mode (MALDI), and capillary voltage, spray voltage, capillary temperature (in the case of ESI) all shown to have dramatic impacts on the quality of MS spectra acquisition. Importantly, there have been recent advances in adding front-end separation technology, such as HPLC or GPC to both ESI and MALDI systems.66–69 The hyphenated systems substantially increase MS analytical power following a reduction in the polymer distribution prior to analysis.
Mass spectrometry is an instrument heavily reliant on calibration to deliver optimum results. Typically, mass is calibrated with standards kits obtainable from a number of commercial providers (external calibration). Time of flight (TOF) mass spectrometers are most prone to error due to TOF variance across the face of the probe plate.24 Standards kits are usually supplied together with the mass spectrometer equipment. Commercial standard kits include Sigma Aldrich's ProteoMassTM, Polymer factory SpheriCal®, and Bruker's starter kit for MALDI-TOF-MS. Commercially available calibration kits are usually mixtures of proteins and peptides with known masses over a certain mass range. For ESI-MS, Thermo Scientifics positive and negative ion calibration solution is available. The calibration solution is a mixture of small molecular weight compounds or peptides with polymers. The limitation of the standards is that they are designed for a certain mass range. Thus, the analyte molecular weight needs to fall within the mass range of the standard to ensure optimal characterization.
Range of Polymers
Over the years, there has been a large range of polymers analyzed via MALDI and ESI MS. Depending on the structure, functionality, and polarity of the polymers, different matrices (for MALDI) and ionization aids (for example salt in ESI) may be required. While sample preparation seems simple, choosing the right experimental conditions and achieving correct sample preparation is critically important in mass spectrometry. Wrong selection of matrices or salt may result in unwanted degradation of the polymers, elimination reactions and in the worst case, failure to “fly” resulting in no MS spectrum. An unwanted by-product of adding an ionization aid like salt is the presence of metal adducts in the MS spectrum. This is due to the ionization process, which proceeds via analyte cation adducts.70
Here, we provide an extensive list of polymers in combination with their recommended MALDI matrices and ESI salt solution formulations taken from previously published studies (Table 4). The exhaustive list of examples is deemed important since there are many variables that affect the result of a mass spectrum such as ionization mode, polymer type, molecular weight, functional groups, matrices choice, salt choice, and solvent. Hopefully, this exhaustive table can provide the reader with reliable information to help in choosing the best analysis conditions for a specific sample analyte.
DEGRADATION INSIDE THE MASS SPECTROMETER
An important matter for consideration when interpreting polymer mass spectra is the degradation or changes that may possibly occur during sample preparation (this is a crucial point when MALDI is used) as well as in the process of transforming the sample from solution or solid (in the case of DESI) to gas phase ions. It is important to remember that even though classified as “soft” ionization, both MALDI and ESI use substantial energy to induce this ionization transformation. In the case of MALDI, the highest risk to polymer degradation/fragmentation is the laser, while for ESI the capillary temperature and the high voltage required to ionize the sample may contribute to some degradation. Another important aspect is induced fragmentation inside the mass spectrometer, which can be very important especially when studying biopolymers using tandem mass spectrometry (MS-MS or MSn).
ESI-MS and MALDI-MS
For an accurate and true interpretation of results obtained using mass spectrometry, it is important to have a clear understanding of what changes may occur to the sample during measurement, via degradation or fragmentation processes, occurring during the ionization phase. It is important to choose the right experimental conditions, matrix and laser energy for MALDI or choice of metal salts and capillary voltage and temperature for ESI. The appropriate experimental conditions for a broad range of polymers have been garnered from the literature and summarized in Table 4. To minimize potential degradation on ionization, the recommended laser strength is set just above the threshold required to generate a spectrum, and the number of laser shots should be minimized as Terrier et al.175 have shown that the number of laser shots used to ionize a given sample in the MALDI analysis of copolymer of poly(ethylene oxide) and poly(propylene oxide) can influence the spectra with an emergence of an unexpected new distribution in the low molecular weight region, as can be seen in Figure 5.
One example of a chemical transformation occurring during the ESI-MS measurement is the incorporation of a metal cation onto a polymer backbone. Work done by Hart-Smith et al.201 and Wesdemiotis and Barner-Kowollik202 showed that poly(acrylic acid), poly(N-iso-propoylacrylamide), poly(diisopropyl vinyl phosphonate), and polyester containing carboxylic acid end groups can be modified by conjugation of the metal counter ion (either Na or K) transforming the proton in R-O-H group into R-O-Na/K during the ionization process in ESI-MS. This work was supported by Giordanengo et al. for poly(methacrylic acid) operated in positive mode, but not in negative mode.203 Moreover, alkoxide formation has also been found to occur during MALDI, as demonstrated by Arakawa et al. and Wyatt et al. for poly(acrylic acid) and poly(methacrylic acid) where it was shown that the addition of salt could be disadvantageous to efficacious MS analysis.50, 63 For poly(ethylene oxide)s with labile end groups, Mazarin et al.72, 204 suggested solvent free sample preparation method with THAP as the matrix. Moreover recent work by our group has shown that extra care needs to be taken when interpreting the MS of polymers bearing labile leaving groups, such as poly(t-butyl methacrylate), where the isopropyl radical can be cleaved forming methacrylic acid groups following fragmentation during ionization in ESI-MS. An analytical area of increasing importance is accurate polymer end-group identification and quantification using MS techniques. Using MALDI, Jackson et al.205, 206 and Boyer et al.207 found that the laser energy could cause fragmentation of bromine terminated PMMA and iodine terminated PMMA resulting in lactone ring formation; in contrast, this transformation was not observed in ESI as shown by Boyer et al.208 Similar chain end modifications during MS was noted by Feldermann et al.209 when analyzing polymers synthesized via RAFT polymerization with dithiobenzoate end groups. The dithiobenzoate group at the chain end was found to either oxidize to form sulfoxides, undergoing substitution of the sulfur with oxygen or fragment via “chugaev” elimination into vinylic end groups depending on the experimental conditions of the ESI-MS, based on analysis of the isotopic pattern; the fragmentation during ionization was much more prominent than disproportion termination, which was observed to be minimal. In other work, MALDI of poly(α-peptoid)s synthesized by ring opening polymerization and initiated by N-heterocyclic carbene (NHC), were shown to undergo NHC elimination in the presence of cationizing salts; in contrast this problem has been shown not to occur in ESI-MS.53, 210 In some rare cases, the ether bond of poly(1,5-dioxepan-2-one) was seen to fragment during the ionization process in ESI; while in contrast fragmentation was not observed in MALDI analysis.167
Tandem Mass Spectrometry (MS/MS) of Homopolymer and Copolymer
Tandem mass spectrometry involves the isolation of a specific ion and the subsequent fragmentation of that ion in a collision cell. MS/MS is a common technique that has been used extensively in proteomic studies for peptide sequencing. However, the utilizations of MS/MS techniques in the field of synthetic polymer analyses have been limited. There are various reasons for this including; complexity of the fragmented spectra making elucidation difficult, as well as experimental challenges. However, the last few years have seen an increased popularity of MS/MS techniques. Notable work by Thalassinos et al. with the development of new software for the assignment of peaks obtained from tandem MS/MS of synthetic polymers.211, 212 MS/MS analyses are able to distinguish isobaric and isomeric species, as well as determine macromolecular connectivities and architecture.59, 170, 202, 213–220 MS/MS information is particularly useful in analyzing degradation mechanisms, quantifying polymer structures, and elucidating monomer sequencing. The range of polymers that can be analyzed includes methacrylates, polyethers, and polystyrenes, with known fragmentation pathways. An example of the software interface is shown in Figure 6.
For further information on MS/MS techniques, we suggest reviews by Crecelius et al. and Wesdemiotis et al.59, 216 Generally, there are two different approaches to tandem mass spectrometry; (i) post source decay (PSD) and (ii) collision induced dissociation (CID).56, 221, 222 PSD is a metastable dissociation of an ion due to either excessive internal energy or via collision with free gas atoms leading to dissociation after the ions are generated in MALDI ionization.56, 223 In CID, fragmentation occurs in the gas phase usually by collision with inert gas molecules (helium, nitrogen, or argon) resulting in bond breakage and fragmentation. These collisions occur in cells located between two mass analyzers. CID (or collisionally activated dissociation, CAD), is the most widely used activation method in polymer analysis.59, 211–213, 220–222, 224–226 Wesdemiotis et al. categorizes the fragmentation of polymer ions into (1) charge-directed polymerization, (2) charge-remote rearrangement, and (3) charge remote fragmentation.216 The CID technique has been used with both MALDI and ESI for the analyses of a wide variety of polymers, including methacrylates, acrylates, poly(ethylene glycol), poly(imide)s, poly(2-ethyl-2-oxasoline)s, poly(sulfone), poly(styrene)s, poly(lactic acids), poly(2-vinyl pyridine), polyesters, polyethers, poly(vinyl acetate), poly(dimethylsiloxane), etc. The fragmentation mode of the polymer ions is dependent on the precursor polymer ions (the location of the charge as well as the type of the charge). Therefore, choosing the right sample preparation conditions is of paramount importance for successful MS/MS analysis (especially the choice of cationizing agent). Mazarin et al.204 have shown that a polymer with nitroxide end groups (containing CON bond) does not fragment in ESI-MS CID if the cationizing agent is a divalent cation such as Cu2+, Zn2+, or Ca2+. In the work of Giordanengo et al.203, poly(methacrylic acid) ions [PMAA+Na]+ and [PMAA-H+2Na]+ ions were shown to fragment via successive dehydration steps as a major pathway.
In another example, CID of polystyrenes [PS + Ag]+ ionized with silver resulted in random homolytic CC bond cleavage generating radical ions which then decomposed by monomer evaporation (primary radical) and backbiting (benzylic radical), as the major degradation pathway (Scheme 1); this result is in accord with the estimation that the weakest bond in polystyrene is the CC bond and molecular modeling of the Ag+ has indicated that the metal should interact with the π electron of the phenyl groups, it is expected that the fragmentation should occur via charge remote CH2CH(Ph) bond scissions.215
A class of polymer that has been studied extensively by tandem mass spectrometry is the poly(ethylene glycol)s. Knop et al.227 investigated MALDI-TOF CID of substituted mPEG with different end groups. CID fragmentation revealed that different end-group dependent fragmentation processes occurred. The most common degradation pathways were 1,4-hydrogen, 1,4-ethylene eliminations, and loss of neutral acids via McLafferty + 1 rearrangements revealing the exact molar mass of the end groups, as can be seen from Figure 7. As shown above, this MS/MS technique can be very useful in end group determination of functionalized polymers.
Tandem mass spectrometry techniques can also be very useful in the study of copolymers, especially for determining the monomer sequence distribution. Wienhofer et al. recently demonstrated that ESI-MS/MS analyses, using an orbitrap MS detector, could provide high resolution mass spectra in high energy decomposition mode, allowing for fragmentation of precursor ions, as well as product ions (which is not possible using CID). The method has shown to be successful in sequencing copolymers of styrene and MMA.182 Similar sequencing was also done by Crecelius et al. to determine the block length of poly(styrene)-b-poly(ethylene oxide) as shown in Figure 8. Baumgaertel et al. studied block copolymers of poly(2-oxazoline), an important class of polymer especially in drug delivery system due to their thermo-responsive properties.220, 228–230 ESI-TOF MS/MS and MALDI TOF MS/MS analyses of various poly(2-oxazoline) block copolymers revealed fragmentation mechanisms such as 1,4-ethylene, hydrogen elimination, and McLafferty +1 rearrangement.228
Another important example of MS/MS utility is sequence and end group determination of complex copolyesters as demonstrated by Weidner et al.231 to differentiate cyclic and linear copolymers by employing MALDI CID MS/MS analyses.
ENVIRONMENTAL POLYMER DEGRADATION
When subjected to a range of environmental conditions, polymers can undergo deteriorative reactions. These conditions include, light, heat, moisture, oxygen, and mechanical stress,.232 Polymer degradation studies have a long history, but recent advances in analysis by mass spectrometry have allowed the quantification of exact changes or modification at the molecular level following degradation. MS is complementary to other instrumentation (IR, UV–vis, GC-MS, SEC, and TGA) that can give information on bulk structural or functional material changes. Soft ionization mass spectrometry provides information on the molecular degradation pathways.71, 233–245 MS can also be used to examine degradation products, for example, in work described by Tondi et al.200 on oligomeric residues present in the cabonisation (900 °C) of polyflavonoid tannin-formaldehyde-furfuryl alcohol or in various HALS degradation studies on polymer coatings.246–249 Moreover, Pastor-Perez et al.168 has also used MALDI to prove that the incorporation of photosensitizer into a poly(glycerol) dendrimer leads to an increase in its photostability. MS techniques can also be used in hyphenation with other instrumentation such as GC-pyrolysis-MS.241, 250–253 In terms of instrumentation, MALDI-TOF is the more common used approach because of its broader mass range. Montaudo et al. and Carroccio et al., have extensively used MALDI-TOF in the investigation of thermal- and photo-oxidation of various polymers including nylon-6,136, 137, 254 nylon-66,138, 255 poly(bisphenol A carbonate),125, 251 poly(ether imide),140, 256 poly(butylene succinate),139, 257 and poly(butylenes terephtalate).148 A specific example illustrating the analytical power of MALDI-MS is the thermal-oxidation degradation of poly(bisphenol A carbonate) bearing different end groups, over a temperature range 300–450 °C. MALDI analyses helped identify an important degradation pathway where hydrolysis of carbonate groups yielded bisphenol A end groups, oxidation of isopropenyl groups, and the oxidative coupling of phenols end groups to form biphenyl groups. The end group of the polymers was found to affect the rate of degradation with degradation products being cyclic oligomers, bisphenol A, and aromatic compounds, such as xanthone, and some insoluble gel indicating cross-linking.125, 251 Coulier et al.258 combined liquid chromatography at critical condition (LC-CC) coupled with MALDI TOF to study the hydrolytic degradation of poly(bisphenol A) carbonate at 100 °C. The combination of LC-MALDI allowed the separation of nondegraded polymer from degraded polymer, using different retention times dependent on end groups, as shown in Figure 9.
In the example of Nylon 6, thermal-oxidation at 250 °C in air, degradation resulted in very complex products with polymer chains identified containing aldehydes, amides, methyl, and N-formamide terminal groups.254 Photo-oxidation of nylon 6 was found to result in oligomers with different end groups following chain cleavage induced by oxidation reactions via mechanisms such as hydrogen abstraction, Norrish I, and Norrish II leading to 40+ different degradation products that could not be identified with other instrumentation.136 Similar degradation products have also been observed following the photo-oxidation of Nylon 66138 while thermal degradation of Nylon 66 at temperatures 290–315°C resulted in gel formation, with hydrolysis products revealed N,N-substituted amide as side chains generated by the condensation of carboxyl end groups with secondary amino groups and azomethine structures originating from the reaction of cyclopentanone moieties with terminal amino groups.255 Degradation mechanisms such as α-hydrogen transfer as well as Norrish I and Norrish II, were found to be important thermal- and photo-oxidation pathways for poly(butylene terephtalate)148 degradation at 280 °C (or under UV light) or poly(butylene succinate)139, 257 (170 °C or under UV light) as studied by Carrococio and coworkers using MALDI-TOF. Similar degradation pathways were also evident in the photo-oxidation of poly(neopentyl isophtalate)145 as observed by Malanowski et al. In some cases, it is important to distinguish between thermal and photo degradation mechanisms, for example 2,2-bis[4-(3,4-dicarboxy-phenoxy)phenyl]propane dianhydride-1,3-phenylendiamine copolymer (ULTEM) degradation140, 256; thermal degradation resulted in (a) cleavage of diphenyl ether units, (b) oxidative degradation of the isopropylidene bridge of BPA units, and (c) thermal cleavage of phenylphthalimide units as depicted in Figure 10. Photo-oxidation, in contrast, resulted in (a) photocleavage of methyl groups of the N-methyl phthalimide terminal units and formation of phthalic anhydride and phthalic acid end groups and (b) photo-oxidative degradation of the isopropylidene bridge of BPA units as depicted in Figure 11.
Another polymer that has been studied extensively is poly(ethylene terephthalate),259 which can undergo discoloration upon degradation. Degradation studies with MALDI-MS of PET in combination with laser desorption/ionization on silicon mass spectrometry (DIOS-MS), has greatly aided the elucidation of the degradation mechanism. Degradation products, terephthalic acid, diethyl terephthalate, and hydroxyl diethyl terephthalate, were identified by Ciolacu et al.147 Blends of PET with polyamide also degraded, with discoloring.184 MALDI-MS also been used to investigate the role of Palladium (Pd) nanoparticles in the accelerated degradation of PET and polyamide (Fig. 12).260
MALDI-TOF-MS has been combined with Py-GC/MS and MALDI TOF/TOF CID techniques, as used by Gies et al. to study changes to poly(p-phenylene sulfide) during processing;162, 252 the additional information obtained by combining these techniques has been highlighted in degradation studies on poly(phenylsulfidesulfone), where good agreement between fragmentation products found in Py-GC/MS and CID indicated that primary cleavage occurred at the phenyl-sulfone bond followed by phenyl-oxygen or phenyl-sulfide bonds as shown in Figure 13.163
To date, the ESI technique has been much less utilized in the study of thermal- or photo-oxidation of polymers, possibly due to its limited mass range. ESI-MS has been used heavily is recent work84, 261–263 on the degradation of poly((meth)acrylates), with different end groups and side chains, under conditions that mimicked the harsh environmental conditions in Australia. Results obtained by ESI-MS suggested that some degradation mechanisms do not precede via radical scission or depolymerization as previously thought, and a degradation mechanism was proposed (based on MS data) as shown in Scheme 2.234, 264–266 The utilization of high resolution ESI-MS, in the form of LTQ-Orbitrap with its enhanced mass accuracy and resolution was critical to this work (Fig. 14).30, 267
In addition to thermal- and photo-oxidation, polymer degradation can occur under ultrasonic exposure, hydrolysis, irradiation, mechanical stress, and chemical/biological exposure. Hoglund and Albertsoon167 investigated residual low molecular weight polymer during the cross-linking of poly(1,5-dioxepan-2-one). Hoskins and Grayson132 studied the hydrolysis of linear and cyclic poly(ε-caprolactone) as shown in Figure 15, while Grayson and Frechet investigated the hydrolytic degradation of aliphatic polyether dendrons.268 In another example, Takeda et al.76, 269, 270 analyzed the ultrasonic degradation of PMMA and other synthetic polymers using MALDI-TOF, and Gunes et al.187 examined the ultrasonic degradation of PET/PEN blends. Dannoux et al.169, 246 used ESI-MS to study the degradation of poly(ether-urethane) under irradiation and Won et al. studied the degradation of poly(ethylene glycol) during a Cu electroplating process.271 Wagner et al.272 used TOF-SIMS to study the degradation of random copolymers of MMA and EGDMA under 5keV SF5+ ion bombardment.
In many of these examples, MALDI or ESI mass spectrometry investigations revealed a vista of new degradation mechanisms. In recent work by Gryn'ova et al.273 revised the mechanism of polymer auto-oxidation. In summary, we believe that mass spectrometry will continue to play an increasing role in probing degradation mechanisms of polymers. Beside ESI and MALDI-MS there are also complimentary techniques employed to study environmental polymer degradation such as Direct probe-atmospheric pressure CI mass spectrometry (DP-APCI) for temporal separation of the thermal degradation products according to the stabilities of the bonds being cleaved274 and DESI mass spectrometry; especially useful for studying changes to polymer surfaces.246, 248
ESI-MS AND MALDI-MS IN CONTROLLED RADICAL POLYMERIZATION
End group modification is an important aspect of modern polymer synthesis, with the increasing popularity of living/controlled free radical polymerization and click chemistry. There are various free radical polymerization techniques that could afford tremendous control over macromolecular design such as: Atom Transfer Radical Polymerization (ATRP),13, 275–281 Nitroxide Mediated Polymerization (NMP),10, 122, 160, 282 Reversible Addition Fragmentation Transfer (RAFT),2, 7, 9, 12, 158, 208, 283, 284 Cobalt Mediated Polymerization,11, 285–289 Cu(0)-mediated living free radical polymerization,290–293 etc. Each polymerization technique is associated with unique end groups. The employment of mass spectrometry (both MALDI and ESI) has helped verify the reaction mechanisms of CLRP techniques, as well as identify any side reactions. In CRP, polymerizations taken above 70% conversion loss of end-group fidelity can cause broadening of MWT distributions. The study of end-groups using mass spectrometry can prove difficult as (by nature) the groups are labile, for example, bromine, dithiobenzoate, trithiocarbonate, and nitroxide, vulnerable to degradation upon ionization especially in MALDI.28, 100, 106, 294 The RAFT polymerization mechanism has been subjected to a number of mass spectrometry studies to probe the mystery of the “intermediate radical,” which was once the subject of much debate.2 Mass spectrometry studies highlighted the susceptibility of dithiobenzoate end groups to cyclic ether solvents such as THF, leading to the recommendation that RAFT polymers should not be stored or dissolved in THF prior to analyses.84 RAFT end group could also degraded via oxidation and “chugaev elimination” in situ MS during ionization resulting in vinylic end group.209 Useful modification of RAFT end group is by aminolysis to form thiol and can further undergo thiol-ene chemistry.295 However, careful condition need to be taken into account since it can also undergo lactone formation as has been investigated by Xu et al.296 Examples of RAFT polymer degradation/modification at end group termini investigated via MS has been summarized in Figure 16.
ATRP polymerization products have also been identified using MS where a loss of end-groups can sometimes lead to proton terminated polymer. This end group loss in ATRP has been shown to occur at high conversions, especially in the presence of excess pentamethyl diethylenetriamine at high conversions, a process known as degradative transfer.276, 297 Hart-Smith et al. used both MALDI-TOF and ESI-MS on star polymers made using ATRP.106 Other free radical polymerization mechanisms, including catalytic chain transfer (CCT) polymerization and self initiated high temperature polymerizations have also been studied using mass spectrometry.298, 299 MALDI-TOF-MS was used to identify end-groups and cobalt-carbon bonding in CCT reactions.300–303 ESI-MS has been used to study the high temperature polymerization of acrylates with different, monomers and solvents.299
Studies on NMP have helped identify nitroxide end group degradation118, 119, 121, 160, 182, 204, 304, 305 Dourges et al.305 used MALDI-TOF to study polystyrenes prepared using 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO). Polystyrene with TEMPO-based alkoxyamine end group was found to undergo gas fragmentation when a combination of dithranol/silver trifluoroacetate (salt) system was used in MALDI-TOF, in contrast with a different matrix (DHB/no salt) where end group fidelity was maintained. Degradation of poly(NIPAAm) TEMPO end groups was observed by Schulte et al.160, 306 Similarly, Bartsch et al.121 studied chlorine-, amine-, and acrylate-functionalized polystyrene capped with TEMPO and found sensitivity to the choice of matrix/salt combination. In studies by Sawamoto et al.281 and Goldbach et al.307 employed dithranol with sodium or silver salts, they noted a loss of chlorine from polystyrene during gas fragmentation or carbon-halogen bond cleavage during sample preparation.281, 307 The fragmentation of end groups in NMP polymerization is highly dependent on the specific alkoxyamine end group, as demonstrated by Dempwolf et al. where alkoxyamines comprised of cyclic structures such as TEMPO, TEMPO-derivatives, and Norpseudopelletierine proved stable during MALDI-TOF analyses while in contrast, N-di-tert-Butyl-O-(1-phenyl-ethyl)-hydroxylamine, N-tert-Butyl-O-(1-phenyl-ethyl)-N-1-isopropyl-2-methyl-propyl)-hydroxyl-amine (BIPNO), and N-tert-Butyl-O-[1-(4-chloromethyl-phenyl)-ethyl]-N-(2-methyl-1-phenyl-propyl)-hydroxylamine (TIPNO) degraded resulting in the loss of t-butyl groups under MALDI-TOF-MS conditions.119, 304 The different fragmentation products have been assigned as shown in Scheme 3. Moreover, polystyrene could also undergo rupture on the backbone resulting in fragments S1 and S2 (Scheme 3).308 Another common NMP agent N-tert-butyl-N-[1-diethoxyphosphoryl]-2,2-dimethylpropyl] nitroxide (SG1) degrades partially when dithranol/silver salt system is employed for analyses but does not show any protonation when DHB/no salt system is used.118 In contrast to MALDI-MS analyses, ESI-MS involves a softer ionization process which does not result in alkoxyamine end group fragmentation, as shown by Delaittre et al. who studied poly(N,N-Diethylacrylamide) with SG1 based end groups.161
Free radical termination mechanisms (disproportionation vs. combination) have been probed in a number of MALDI and ESI MS studies.290, 293, 309, 310 Recently, MS studies have focused on Cu(0) mediated living radical polymerizations,290, 311 using ESI-MS, NMR, and GPC in combination, showing that transfer reactions to ligand (resulting in species 3) and disproportionation (resulting in species 2 and 3 in equal ratio). ESI-MS has allowed the identification of main chain termination reactions during the polymerization of methyl acrylate via Cu(0) mediated polymerization: in one study, the bromine terminated polymer undergoes H-abstraction presumably with the ligand to give a nonfunctional proton terminated polymer as shown by Species 3 in Figure 17. Interestingly, ESI has also indicated the absence of coupled polymer chains. The ESI-MS analyses of copper mediated polymerizations were heavily aided by high resolution LTQ-Orbitrap MS operated in FTMS mode, enabling resolution of overlapping isotopic patterns as shown in Figure 17.
It is important to judge the performance of mass spectrometry in comparison to other analytical techniques standard in polymer characterization such as NMR spectroscopy. In particular, characterizing end groups or other functional groups in a low molecular weight polymer such as those synthesized via CRP. There are some important parameters, which make MS complement or even supersede NMR. Generally, in NMR analyses, identification of a specific group is dependent on it's quantity ratio with another functionality—making it harder to analyze as the molecular weight of the polymer increases, especially if the signal peaks overlap with other signal from the polymer backbone. However, in mass spectrometry this is not a problem as the mass difference is discrete. Mass spectrometry is also a powerful analytical method to use in the verification of post-polymer modification (obviating the need for purification since small molecules do not interfere with the mass spectral analysis of the polymer). Most importantly mass spectrometry gives molecular information while NMR gives structural information. For example, NMR characterization of Cu(0) mediated polymerization will show the loss of end groups while MS provides mechanistic information, namely, that end group loss can be attributed to chain transfer and disproportionation as shown by Nystrom et al.311 Another example of the power of mass spectrometry is the study of products from polymer degradation during polymerization or postmodification reactions, often difficult to purify from final products, for example, degradation of RAFT end groups during storage,84 or end group modifications on ATRP synthesized polymers.312 Tandem MS/MS is powerful for not only functional group identification but also location and bond strength. Examples of fragmentation MS/MS have been summarized in an informative review by Crecelius et al.59
ESI-MS AND MALDI-MS IN POLYMER ARCHITECTURE
Another area of polymer science where mass spectrometry has found application is in the characterization of complex macromolecular architecture such as star polymers, dendrimers, and cyclic polymers. Characterization of such complex structures can be problematic as the products may be very similar in nature to starting materials and/or by-products. Dendrimers are highly branched 3D structures with multi-functional groups, of highly defined molecular weight and size. Dendritic molecular structures can be synthesized via divergent or convergent methods.313, 314 Since the seminal work Hawker and Frechet in the 1990s' on the convergent growth approach to dendrimers, mass spectrometry has been an integral characterization technique often performed with EI and FAB,315 as at the time “soft” mass spectrometry unit analysis was still in preliminary development. However, later work on dendrimers made use of modern mass spectrometry analysis especially MALDI. Jun et al.316 studied PEGylated drug conjugates on poly(arylether) dendrimers with MALDI-TOF-MS as the main characterization approach. MALDI analysis by Yu et al.317 of various dendrons was used to differentiate between AB dendritic-linear diblocks, ABA triblocks, and linear PEG by end group analysis; in this work, MALDI also provided more accurate molecular weight data then GPC even in the linear mode. MALDI-TOF is also useful for characterizing dendritic material comprised of conducting polymer as shown by Wiesler et al.318 with dendrimers of polyphenylene and dendritic polythiophene by Malenfant and Frechet.319 Zhou et al.320 used MALDI-TOF creatively not only to determine the structure and polydispersity of dendrimers but also to quantitatively analyze the number of metal Cu2+ ions that could be complexed by the dendrimer. One efficient and effective way to synthesis dendrimer structures is by utilizing “click” chemistry as demonstrated in work by Wu et al.321 and Lee et al.322 using Azide Alkyne Huisgen cycloaddition to form dendrimers via both convergent and divergent routes. Another powerful click technique that has been utilized especially by Hawker and coworkers is thiol-ene coupling.323, 324 The combination of Thiol-ene and CuAAC “click” has also been used to synthesize sixth generation dendrimers in a single day.325 MALDI-TOF-MS has proved to be a vital characterization technique in all of these studies. Another class of polymer architecture difficult to characterize is cyclic polymer. Most cyclic polymers are prepared from precursor linear polymer. Soft ionization mass spectrometry has greatly assisted in characterizing cyclic chains as shown by Quirk et al.326 in the synthesis of cylic polystyrene via living anionic polymerization and metathesis ring closure. The structure of cyclic polymer was evidenced by a characteristic loss of molecular weight distinguishing the cyclic form from its linear precursor. In the case of cyclic structure formation in the absence of mass change (cyclic and linear polymeric isomers) as in the case for cyclic polymer prepared via CuAAC chemistry, mass spectral characterization is more limited. Hoskins et al.327 employed ion mobility spectrometry (IMS) to separates ions according to their size in the gas phase, coupled with MALDI to correlate the mass to charge ratio (m/z) with drift time (related to size). Li et al.189, 328 also combined IMS with both MALDI and ESI as well as tandem mass spectrometry to analyze the complex synthesis of poly(α-peptoid)s and subsequent supramolecular structure formation.
The development of soft ionization mass spectrometric techniques such as MALDI and ESI, has revolutionized the molecular characterization of macromolecules enabling data acquisition yielding composition at the single chain scale. MALDI and ESI analyses have increased understanding of mechanisms of environmental polymer degradation, enzymatic polymer degradation, chemical polymer degradation, mechanistic studies of various polymerizations (especially Controlled/Living Free Radical Polymerization), and chemical(click) modification of polymers, as outlined in this review.
Soft ionization mass spectrometry is mostly used in proteomics and protein structural studies, and so general methodologies have largely derived from studies on biomolecules. In comparison with structural databases on proteins and peptides, data on polymers and their fragmentation patterns are sparse, and improvements in software databases to accompany instrumentation will certainly enhance the value of mass spectrometry instrumentation in general Polym. Chem. laboratories. Recent developments in ion mobility mass spectrometry have demonstrated a capacity for probing polymer architecture and supramolecular structures.189 The uptake of ESI instrumentation has been slower in the polymer research community, as more expertise is required in maintaining and efficiently running the instrumentation; however, ESI has tremendous potential as ionization processes tend to be “softer” than MALDI, an important issue for preserving labile end-groups. In addition, ESI can be coupled with SEC.28 The use of advanced MS techniques in polymer science is still in its infancy, but the examples given in this review, indicate the power of MS analytical approaches for probing complex molecular mechanisms by allowing single chain analyses, this in turn leads to increased mechanistic understanding providing important information in designing new efficient polymerization reactions.
The authors would like to thank all their colleagues, collaborators, and students who have contributed to work on mass spectrometric analyses of polymers at UNSW. Dave Haddleton sparked initial interest during a collaborative study with TPD on MALDI studies on free radical termination and propagation mechanisms, and CCT in the mid-1990s. They acknowledge UNSW capital funding for the funds to purchase an ESI mass spectrometer in 1999 (to T.P. Davis) that significant expanded their capacity to work on living/controlled polymerization (especially RAFT) and polymer degradation in recent years. They wish to acknowledge specifically Christopher Barner-Kowollik and Phil Vana who both contributed significantly to mass spectrometry research during post-doctoral studies with T.P. Davis at UNSW and who have both gone on to become leading advocates of MS characterization in polymer science research. C. Boyer is thankful for his ARC-APD (DP1092640) and Future Fellowship (FT120100096).
Alexander H. Soeriyadi, born in Indonesia, studied chemical engineering at University of New South Wales receiving B.Eng (Hons Class 1) in 2008. He undertook his PhD studies in CAMD group under Prof. Thomas P. Davis, graduating in 2012. He is currently a Research associate in Australian Centre of Nanomedicine (ACN) working with Prof. Justin Gooding. His research interests are stimuli responsive material and its applications, living polymerization techniques, and mass spectrometry in polymer science.
Michael Whittaker is currently a Senior Research Fellow and the Research Manager for both the Centre of Advanced Macromolecular Design (CAMD) and the Australian Centre for Nanomedicine (ACN), University of New South Wales. He has coauthored over 70 research papers and patents. His research interests include; toxicology of nanoparticles, novel applications of “click” chemistry in polymer synthesis, synthesis of “smart” hybrid inorganic/organic nanomaterials for nanomedicine and new applications of polymerization methods in polymer design.
Cyrille Boyer received his PhD in polymer chemistry from the University of Montpellier II. His PhD was in collaboration with Solvay-Solexis and devoted at the synthesis of new graft copolymers using grafting “onto.” At the end of his PhD, he undertook an engineer position with Dupont Performance and Elastomers, dealing with the synthesis of original fluorinated elastomers. Later, he joined the Centre for Advanced Macromolecular Design (CAMD) as a senior research fellow. He was working on the synthesis of protein-polymer conjugates using RAFT technology. In 2010, he has been appointed as a lecturer in the School of Chemical Engineering and he received an Australian Research Council Fellow (APD-ARC). In 2011, he joined the Australian Centre for NanoMedicine as a project leader to develop new polymeric nanoparticles for drug delivery. The same year, Cyrille started a new research program on the preparation of new hybrid nanoparticles/polymers for hydrogen storage in collaboration with Francois Aguey-Zinsou (School of Chemical Engineering). In 2012, Cyrille has been appointed Senior Lecturer at the School of Chemical Engineering and he has been awarded an ARC-Future Fellowship. Cyrille's research interests mainly cover the preparation of well-defined polymeric nanoparticles for drug delivery and siRNA delivery, protein polymer conjugates, hybrid organic-inorganic nanoparticles for imaging and energy storage, and macromolecular design. Cyrille has co-authored over 90 research articles, 2 book chapters and 3 international patents.
Tom Davis has been an academic at UNSW for 17 years following a stint in industry as a research manager at ICI in the United Kingdom. He has coauthored 325+ refereed papers, patents, and book chapters. He is the founding Director of the Centre for Advanced Macromolecular Design (CAMD) at UNSW–where he initiated and built up significant research programs in free radical polymerisation (pulsed-laser polymerisation); copolymerisation mechanism; catalytic chain transfer, RAFT kinetics, mechanism and synthesis of complex architectures (stars, microgels, blocks etc); honeycomb films from breath figures; glycopolymers and enzymatic synthesis; biodegradable polymers; soft ionisation mass spectrometry of polymers (MALDI and ESI); polymer–protein hybrids and hybrid nanoparticles. He is also a visiting Professor at the Institute for Materials Research & Engineering (IMRE) in Singapore. In 2005, he was awarded a Federation Fellowship by the Australian Research Council. He serves (or has served) on the editorial advisory boards of Macromolecules, Journal of Polymer Science, Australian Journal of Chemistry, Journal of Materials Chemistry,Journal of Macromolecular Science – Reviews and Polymer Chemistry.