Anal. Chem. 2007, 79, 4814-4819
Principle of Two-Dimensional Characterization of Copolymers Steffen Weidner,* Jana Falkenhagen, Ralph-Peter Krueger,† and Ulrich Just
Federal Institute for Materials Research and Testing (BAM), Department I.3, Richard-Willstaetter-Strasse 11, D-12489 Berlin, Germany
Two-dimensional polymer characterization is used for a simultaneous analysis of molar masses and chemical heterogeneities (e.g., end groups, copolymer composition, etc.). This principle is based on coupling of two different chromatographic modes. Liquid adsorption chromatography at critical conditions (LACCC) is applied for a separation according to the chemical heterogeneity, whereas in the second-dimension fractions are analyzed with regard to their molar mass distribution by means of size exclusion chromatography (SEC). Because appropriate standards for a calibration of the SEC are seldom available, matrix-assisted laser desorption/ionizationtime-of-flight mass spectrometry (MALDI-TOF MS) was used to substitute the SEC. The LACCC-MALDI MS coupling enables acquiring additional structural information on copolymer composition, which can considerably enhance the performance of this coupled method. Misleadingly, liquid chromatography of polymers is often associated with size exclusion chromatography (SEC). Polymer molecules are dissolved in a thermodynamically “good” solvent in that particular mode of chromatography. In an ideal case, there are no interactions between the stationary phase and dissolved molecules (∆H ) 0). The separation is based on entropic parameters (exclusion) according to the hydrodynamic volume of dissolved polymer molecules. To determine the exact molar masses, every SEC system has to be calibrated using standards with the same structure as that of the investigated polymer. But even the slightest modifications of the polymer structure (variation of end groups, tacticity, etc.) can cause dramatic changes in exclusion behavior. In addition to changes in the hydrodynamic volume of dissolved polymers, interactions between molecules and stationary phase may occur. These interactions are used for a separation in liquid adsorption chromatography (LAC). Thermodynamically, this can be expressed through ∆H < 0 and T∆S , ∆H. Due to strong differences in the polarity of the stationary phase and polymer molecules, LAC is often not applicable for polymer chromatography. Mostly, a strong adsorption of end groups and groups located in the repeat units can be observed, which makes this technique only feasible for a separation of specific oligomers and low-mass additives. * Corresponding author. E-mail:
[email protected]. Fax: +49 30 81041337. † In memoriam Dr. Ralph-Peter Krueger, December 22, 1947-April 2, 2007.
4814 Analytical Chemistry, Vol. 79, No. 13, July 1, 2007
A balance of entropic and enthalpic interactions can be achieved at the so-called “critical point” of adsorption using solvent mixtures. A thermodynamic expression for the free enthalpy is given by the Gibbs-Helmholtz equation
∆G ) ∆H - T∆S This extraordinary behavior is characterized by an exclusive interaction of single functional groups (e.g., end groups) with the stationary phase. In this special case, polymer molecules elute according to end group polarity independently of their molar masses. But the principle of “critical” separation can be applied to many other polymer heterogeneities (chirality, tacticity, block length of copolymer blocks).1-21 For a two-dimensional polymer characterization, the two modes are combined in one chromatographic system. Fractions are extracted from the first dimension by means of a valve and (1) Adrian, J.; Braun, D.; Rode, K.; Pasch, H. Angew. Makromol. Chem. 1999, 267, 73-81. (2) Biela, T.; Duda, A.; Rode, K.; Pasch, H. Polymer 2003, 44, 1851-1860. (3) Braun, D.; Esser, E.; Pasch, H. Int. J. Polym. Anal. Charact. 1998, 4, 501+. (4) Falkenhagen, J.; Friedrich, J. F.; Schulz, G.; Kruger, R. P.; Much, H.; Weidner, S. Int. J. Polym. Anal. Charact. 2000, 5, 549-562. (5) Falkenhagen, J.; Much, H.; Stauf, W.; Muller, A. H. E. Macromolecules 2000, 33, 3687-3693. (6) Falkenhagen, J.; Weidner, S. M. Rapid Commun. Mass Spectrom. 2005, 19, 3724-3730. (7) Gorshkov, A. V.; Much, H.; Becker, H.; Pasch, H.; Evreinov, V. V.; Entelis, S. G. J. Chromatogr. 1990, 523, 91-102. (8) Gorskov, A. V.; Jevreinov, V. V.; Lausecker, B.; Pasch, H.; Becker, H.; Wagner, G. Acta Polym. 1986, 37, 740-741. (9) Heinz, L. C.; Macko, T.; Pasch, H.; Weiser, M. S.; Mulhaupt, R. Int. J. Polym. Anal. Charact. 2006, 11, 47-55. (10) Keil, C.; Esser, E.; Pasch, H. Macromol. Mater. Eng. 2001, 286, 161-167. (11) Kruger, R. P.; Much, H.; Schulz, G.; Rikowski, E. Monatsh. Chem. 1999, 130, 163-174. (12) Pasch, H. Macromol. Symp. 1996, 110, 107-120. (13) Pasch, H.; Augenstein, M.; Trathnigg, B. Macromol. Chem. Phys. 1994, 195, 743-750. (14) Pasch, H.; Brinkmann, C.; Gallot, Y. Polymer 1993, 34, 4100-4104. (15) Pasch, H.; Brinkmann, C.; Much, H.; Just, U. J. Chromatogr. 1992, 623, 315-322. (16) Pasch, H.; Deffieux, A.; Henze, I.; Schappacher, M.; RiqueLurbet, L. Macromolecules 1996, 29, 8776-8782. (17) Pasch, H.; Esser, E.; Kloninger, C.; Iatrou, H.; Hadjichristidis, N. Macromol. Chem. Phys. 2001, 202, 1424-1429. (18) Pasch, H.; Rode, K. Polymer 1998, 39, 6377-6383. (19) Pasch, H.; Rode, K.; Chaumien, N. Polymer 1996, 37, 4079-4083. (20) Trathnigg, B.; Thamer, D.; Yan, X.; Maier, B.; Holzbauer, H. R.; Much, H. J. Chromatogr., A 1994, 665, 47-53. (21) Wachsen, O.; Reichert, K. H.; Kruger, R. P.; Much, H.; Schulz, G. Polym. Degrad. Stab. 1997, 55, 225-231. 10.1021/ac062145f CCC: $37.00
© 2007 American Chemical Society Published on Web 06/02/2007
Figure 1. Reaction scheme, structure, and molar masses of investigated copolyesters (AS, adipinic acid; NPG, neopentylglycol; HD, hexanediol).
transferred into the second dimension. Usually, liquid adsorption chromatography at critical conditions (LACCC) is used for a chemical separation in the first dimension, whereas size exclusion chromatography (SEC) gives additional mass information in the second dimension. The scheme of a 2D chromatography (also referred to as “orthogonal liquid chromatography”) and various applications of this technique can be found in the literature.22-31 However, the coupling of two chromatographic systems operating with different solvents and salts frequently causes problems; e.g., when transferring fractions from the LACCC to the SEC solvent or salt, signals can overlay polymer signals in the lower molar mass region. But the main disadvantage is the lack of appropriate standards for calibrating the SEC system. Hence, the advantage of the functionality type separation in LACCC might be neutralized by the wrong masses. In order to overcome these disadvantages, matrix-assisted laser desorption/ionization-time-of-flight mass spectrometry (MALDI-TOF MS) was used to substitute SEC. The calibration of a MALDI-TOF mass spectrometer can be performed by using any suitable substance, e.g., biopolymers or synthetic polymers. But MALDI-TOF MS also has some drawbacks. (22) Adrian, J.; Braun, D.; Pasch, H. Angew. Makromol. Chem. 1999, 267, 8288. (23) Adrian, J.; Esser, E.; Hellmann, G.; Pasch, H. Polymer 2000, 41, 24392449. (24) Biela, T.; Duda, A.; Penczek, S.; Rode, K.; Pasch, H. J. Polym. Sci., Part A: Polym. Chem. 2002, 40, 2884-2887. (25) Falkenhagen, J.; Much, H.; Stauf, W.; Muller, A. H. E. Abstr. Pap. Am. Chem. Soc. 1999, 218, U488-U488. (26) Heinz, L. C.; Siewing, A.; Pasch, H. E-Polymers 2003. (27) Kruger, R. P.; Much, H.; Schulz, G. J. Liq. Chromatogr. 1994, 17, 30693090. (28) Kruger, R. P.; Much, H.; Schulz, G.; Wachsen, O. Macromol. Symp. 1996, 110, 155-176. (29) Pasch, H. Phys. Chem. Chem. Phys. 1999, 1, 3879-3890. (30) Pasch, H. Macromol. Symp. 2001, 174, 403-412. (31) Siewing, A.; Schierholz, J.; Braun, D.; Hellman, G.; Pasch, H. Macromol. Chem. Phys. 2001, 202, 2890-2894.
Polymer samples have to exhibit a narrow molar mass distribution of maximum 1.2. Most industrial polymers, however, are broadly distributed materials. Thus, the separation in LACCC mode, which results in sharp polymer peaks separated according to end groups, has to be modified in order to obtain an additional mass separation. This can be done by shifting the LACCC conditions either into SEC or, better, into LAC mode. As a result, the separation according to functionalities (LACCC) is overlaid by a molar mass separation. A fractionation of those samples for MALDI-TOF MS results in narrow (mass) distributed polymer fractions with definite end groups. Due to its superior mass resolution, ESI-TOF mass spectrometry was applied for an accurate end group determination. In contrast to SEC, which only yields molar masses and mass distributions, MALDI-TOF MS provides additional information on polymer structures. This could be valuable, especially for determining the composition of copolymers, which can be regarded as a supplementary third dimension of polymer analysis. This paper demonstrates how useful the substitution of SEC by MALDI-TOF MS is using the complex analysis of a polyester copolymer as an example. EXPERIMENTAL SECTION Samples. The polyester samples have been provided by Bayer Co. (Leverkusen, Germany). Their structures are shown in Figure 1. Samples were synthesized using an excess of diols. Therefore, the preferred formation of polymers with hydroxyl end groups was expected. LACCC. A Hewlett-Packard HP 1090 liquid chromatograph equipped with two Silica gel columns (300 + 100) Å with 5-µm pore size, 250 + 125 mm length, and 4.6-mm i.d. (MachereyNagel) were applied for chromatographic separation at the “critical” point of adsorption. An evaporative light scattering Analytical Chemistry, Vol. 79, No. 13, July 1, 2007
4815
Figure 2. Chromatogram of a copolymer recorded at the critical point of adsorption of adipinic acid (acetone/n-hexane 48/52 (v/v), silica gel columns, 45 °C, 0.5 mL/min, ELSD).
Figure 3. 2D plot combining the structural information of separation at LACCC mode (Figure 2) and molar mass information of 100 SEC chromatograms.
detector (ELSD) SEDEX 45 (Sedere) was used for detection. The chromatographic system was kept at a constant temperature of 45 °C. The eluent flow was adjusted at 0.5 mL/min. An approximate “critical solvent composition” was adjusted using a solvent system of acetone/n-hexane 48/52 (v/v). MALDI-TOF MS. A Reflex III MALDI mass spectrometer (Bruker-Daltonik), operating at 20-kV acceleration voltage, was used. Ionization/desorption was performed by a UV laser working at a wavelength of 337 nm. The laser pulse length was 3 ns. Typically, 100-200 transients were accumulated for one spectrum. Preferably, sodium adduct ions were formed during ionization. 2-[(2E)-3-(4-tert-Butylphenyl)-2-methylprop-2-enylidene]malononitrile (DCTB) (Fluka) was used as a matrix. The mass spectrometer was calibrated using poly(ethylene oxide) standards with an average molar mass of 2000 and 6000 g/mol. ESI-TOF MS. The experiments were performed using a Q-TOF Ultima API (Waters, Eschborn, Germany) hybrid mass spectrometer equipped with an orthogonal electrospray source (Z-spray) operating at 3 kV and controlled by MassLynx 4.0 software. The source temperature was 120 °C; the desolvation temperature was 250 °C. The TOF mass spectrometer operated at an acceleration voltage of 9.1 kV, a cone voltage of 50 V, 4816 Analytical Chemistry, Vol. 79, No. 13, July 1, 2007
Figure 4. MALDI-TOF mass spectra of peaks taken after fractionation of an LACCC run (see Figure 2 at (1) representing the peak at 8.5 min, (2) at 9.6 min, (3) at 10.4 min, and (4) at 11.4 min of retention time) and enlargement showing the pattern of obtained peak clusters and their mass difference.
Figure 5. Chromatogram of monomers recorded at the critical point of adsorption of adipinic acid (acetone/n-hexane 48/52 (v/v), silica gel columns, 45 °C, 0.5 mL/min, ELSD).
and a collision energy of 10 eV. The samples were introduced by means of a syringe pump. The mass range up to m/z ) 1500 was continuously scanned at a rate of one spectrum per second. Coupling of LACCC with MALDI-TOF-MS. A commercially available interface LC 500 (LabConnections) was used for semion-line coupling of liquid chromatography and mass spectrometry. The MALDI target was precoated with matrix before fractionation of the polymer. The matrix DCTB, dissolved in THF (∼1 mg/ mL), was sprayed by means of a secondary pump (Knauer). The flow rate of the added matrix solution was 0.2 mL/min. The flow was focused onto the MALDI target. The solvent was evaporated in a nitrogen gas flow at an elevated temperature of 120 °C. The temperature must be carefully adjusted to avoid crystallization of matrix on the tip. The transfer system was controlled by software that enables the automatic assignment of sample spots to the corresponding retention times. RESULTS Figure 2 shows the chromatogram recorded near the “critical” point of adsorption of adipinic acid (AS). Four peaks can be seen at retention times of 8.4, 9.6, 10.5, and 11.4 min. This corresponds
Figure 6. ESI-TOF mass spectra of copolymer structures representing peaks 1 (a) and 4 (b) of the chromatogram shown in Figure 2.
Figure 7. Series of 9 MALDI-TOF mass spectra recorded using fractions obtained by spraying the LACCC run (shown in Figure 2) from 8 to 12 min of retention time.
very well to the assumed formation of four different speciess rings and copolymers with either dineopentyl end groups (di-NPG) or dihexanole end groups (di-HD) or mixed neopentyl/hexanole end groups (NPG/HD). The coupling of LACCC with SEC in the second chromatographic dimension results in over 100 SEC chromatograms. The run time of one single SEC analysis took only 2 min. The resulting 2D plot is shown in Figure 3. In addition to the separation at critical conditions, this plot provides additional information on the molar mass at each point of the LACCC chromatogram and on the concentration of the
Figure 8. 2D plot combining the structural information of separation at LACCC mode (Figure 2) and molar mass information of 12 MALDITOF mass spectra recorded after spraying the LACCC run continuously onto the MALDI target.
different species. However, using this coupling method, neither structures can be revealed nor exact molar masses can be obtained due to the lack of suitable standards for SEC calibration. For this reason, MALDI TOF mass spectrometry was used. The corresponding MALDI-TOF mass spectra of manually fractionated Analytical Chemistry, Vol. 79, No. 13, July 1, 2007
4817
Figure 9. Copolymer composition depending on the end groups (mass numbers taken from MALDI and intensity from chromatography).
LACCC peaks are illustrated in Figure 4. The shape of these spectra is typical for polymers with a broad molar mass distribution. Very intense signals around 1.000 m/z could be observed, followed by a long tailing toward greater masses. In the enlargement groups of peaks with differences of 14 Da within each cluster can be seen. This mass difference can be explained by a replacement of [NPG-AS] units by [HD-AS] units. By means of MALDI-TOF MS, an assignment of structures to assumed end group was not possible, because the mass difference of the end groups (14 Da) overlaps with the mass difference of the repeat units, which is also 14 Da. Thus, it was crucial to know the structure of end groups for an accurate determination of the copolymer composition. As mentioned earlier, the separation in LACCC is based on interactions of end groups with the stationary phase. Therefore, “monomers” (HD, neopentyl glycol (NPG), and AS) were first used for a separation at the chosen “critical” conditions. Their elution order is shown in Figure 5. The chromatogram clearly reveals the different retention times of NPG (5.8 mL) and HD (7.2 mL). A similar order can be expected for the elution of polymers at critical conditions of adsorption. A reliable determination of end groups could be achieved by means of Electrospray-time-of-flight mass spectrometry. This method offers very high mass resolution especially in lower mass regions. For this purpose, fractions of the chromatogram shown in Figure 2 were extracted. The ESI-TOF mass spectra corresponding to peaks 1 and 4 are shown in Figure 6a + b. The existence of assumed cyclic structures (Figure 6a) could be clearly proved, whereas peaks shown in Figure 6b could be attributed to different 4818
Analytical Chemistry, Vol. 79, No. 13, July 1, 2007
copolymer compositions, depending on end group structures (HD or NPG). However, a closer look at the spectra shows that, assuming NPG end groups, some of the peaks cannot be assigned to an appropriate copolymer structure. In contrast to that, an assumption of HD end groups provided reasonable data for all peaks. The formation of additional K+ ions can only be observed for the last peak of each Na+ adduct series, whereas the other K+ peaks (+16 Da) are superimposed by Na+ peaks showing a difference of +14 Da to each other. The enlargement of the isotopic patterns shows irregularities indicating a superposition of two distributions having a difference of 2 Da. The final clarification was obtained from the elemental composition report, which unambiguously shows that the peak at 1039.6176 Da can only be represented by a molecule with a sum formula of C53H92O18Na1, which can be attributed to an [A3B1] copolymer with HD end groups. This result is in accordance with the previously shown elution order of the “monomers” (see Figure 5). In order to compare ordinary 2D chromatography (LACCC-SEC) with the new LACCC-MALDI TOF MS principle, the chromatographic run at critical conditions of adsorption was transferred by spraying onto the MALDI target. A series of these raw spectra, recorded at corresponding retention times between 9 and 12 min (see Figure 2) is shown in Figure 7. In contrast to the MALDI spectra of the manually fractionated peaks (Figure 4), baseline-resolved series of peaks could be obtained up to a mass range of 10.000 g/mol. A total of 12 MALDI TOF mass spectra were recorded, normalized, and combined with intensities extracted from chromatography. These data were used to create a 2D plot (see Figure
8). A comparison with Figure 3 clearly shows similarities even in the details. But in contrast to the plot based on chromatography, the 2D plot obtained by LACCC-MALDI coupling offers a more reliable mass scale. Knowing the end group structure, the exact copolymer composition can be determined. For that purpose, a homemade program was used to construct a matrix composed of mass numbers representing potential block compositions, depending on the end group mass. Each MALDI mass spectrum was scanned for mass numbers contained in the matrix. In the event of a positive match, the chromatographically corrected peak intensity was added. Finally, the matrix was used to create a 2D composition plot providing the intensity information for any existing copolymer composition depending on the end group structure. These 2D composition plots are shown in Figure 9. Surprisingly, the copolymer composition differs from theoretical values. As reported by another group32 and expected by the mechanism of condensation, a nearly 1:1 ratio of co-monomer integration was supposed. However, the 2D plots clearly indicate a different behavior independently of the nature of the end group, which results in a higher content of [HD-AS] units. A possible (32) Willemse, R. X. E.; Ming, W.; van Herk, A. M. Macromolecules 2005, 38, 6876-6881.
explanation can be given either by various ionization probabilities of the repeat units in the MALDI-TOF process or by a different reactivity of diols. SUMMARY A new principle of 2D analysis of polymers is described based on the coupling of liquid adsorption chromatography at the critical point of adsorption with MALDI TOF mass spectrometry. LACCC enabled the separation of polymers according to their end group structure. The substitution of SEC by MALDI TOF MS offers considerable advantages, e.g., no need for a polymer-dependent calibration. In addition, it has been demonstrated that, after separation and fractionation, MALDI TOF MS can reveal the composition of complex copolymers. ACKNOWLEDGMENT The authors thank Dr. Schweer (Bayer Company) for providing the polymer samples. Received for review November 14, 2006. Accepted April 30, 2007. AC062145F
Analytical Chemistry, Vol. 79, No. 13, July 1, 2007
4819
