Anal. Chem. 1998, 70, 131-135
Use of MALDI-TOF To Measure Molecular Weight Distributions of Polydisperse Poly(methyl methacrylate) Hassan Rashidzadeh and Baochuan Guo*
Department of Chemistry, Cleveland State University, Cleveland, Ohio 44115
The main problem encountered in the MALDI-TOF analysis of polydisperse polymers is mass discrimination against high-mass oligomers. This work investigated some of the causes of this problem by using PMMAs as the polymer analytes. It was found that both instrumental and matrix factors could lead to this problem. Among the instrumental factors, detector saturation resulting from strong signals of matrix-related and low-mass oligomer ions can be a potential major cause of this problem. Since most of the ion detectors do not have an adequate dynamic range to avoid saturation, detection saturation could be a fundamental limitation, especially when the molar ratio of high- to low-mass oligomers is small. A quantitative analysis was also performed to examine the matrix effect. IAA and HABA were selected for this study. It was found that mass discrimination occurred in both cases, but the use of HABA led to more profound mass discrimination. This shows that the use of improper matrixes could be another source causing mass discrimination. Hence, unless new approaches are developed, one must be cautious in using MALDI-TOF for directly measuring MWDs of polydisperse polymers, especially those highly polydisperse polymers. Matrix-assisted laser desorption/ionization time-of-flight (MALDITOF) mass spectrometry is a recently introduced method for characterization of high molecular weight molecules. One of its potential applications is measurement of molecular weight distributions (MWDs) of synthetic polymers.1 In the past several years, a number of research groups have actively examined the use of MALDI-TOF for polymer analysis.2-28 The result reveals that the (1) Bahr, U.; Deppe, A.; Karas, M.; Hillenkamp, F. Anal. Chem. 1992, 64, 2866. (2) Danis, P. O.; Karr, D. E. Org. Mass. Spectrom. 1993, 28, 923. (3) Danis, P. O.; Karr, D. E.; Holle, A.; Waston, C. H. Org. Mass. Spectrom. 1992, 27, 843. (4) Juhasz, P.; Costello, C. E.; Biemann, K. J. Am. Soc. Mass Spectrom. 1993, 4, 399. (5) Bu ¨ rger, H. M.; Mu ¨ ller H. M.; Seebach, D.; Bo ¨rnsen, K. O.; Scha¨r, M.; Widmer, H. M. Macromolecules 1993, 26, 4783. (6) Danis, P. O.; Karr, D. E.; Simonsick, W. J.; Wu, D. T. Macromolecules 1995, 28, 1229. (7) Tang, X.; Dreifuss, P. A.; Vertes, A. Rapid Commun. Mass Spectrom. 1995, 9, 1141. (8) Williams, J. B.; Gusev, A. I.; Herciles, D. M. Macromolecules 1996, 29, 8144. (9) Jackson, C.; Larsen, B.; McEwen, C. Anal. Chem. 1996, 68, 1303. (10) Larsen, B. S.; Simonsick, W. J., Jr.; McEwen, C. J. Am. Soc. Mass Spectrom. 1996, 7, 287. S0003-2700(97)00568-4 CCC: $14.00 Published on Web 01/01/1998
© 1997 American Chemical Society
MWD data obtained from MALDI-TOF can be in agreement with those determined from conventional methods such as gel permeation chromatography (GPC) if narrowly distributed polymers are analyzed and the sample tips are carefully prepared.27,28 However, the currently used MALDI-TOF method has failed to provide reliable MWDs of polydisperse polymers. To use MALDI-TOF for measurements of MWDs of polydisperse polymers, an indirect approach involving a combination of MALDITOF and GPC has been proposed.17,20 In this method, a polydisperse polymer is fractioned by using GPC, yielding narrow distribution fractions followed by the MALDI-TOF analysis of each of the fractioned samples. But, in this way, MALDI-TOF is being used as a mass-selective detector. As has been suggested,19 the GPC calibration by MALDI-TOF is certainly a useful procedure to obtain reliable MWDs of polydisperse polymers and this (11) Lehrle, R. S.; Sarson, D. S. Rapid Commun. Mass Spectrom. 1995, 9, 91. (12) Belu, A. M.; DeSimone, J. M.; Linton, R. W.; Lange, G. W.; Friedman, R. M. J. Am. Soc. Mass Spectrom. 1996, 7, 11. (13) Montaudo, G.; Montaudo, M. S.; Puglisi, C.; Samperi, F. Anal. Chem. 1994, 66, 4366. (14) Montaudo, G.; Montaudo, M. S.; Puglisi, C.; Samperi, F. Rapid Commun. Mass Spectrom. 1994, 8, 981. (15) Montaudo, G.; Montaudo, M. S.; Puglisi, C.; Samperi, F. Rapid Commun. Mass Spectrom. 1994, 8, 1011. (16) Montaudo, G.; Montaudo, M. S.; Puglisi, C.; Samperi, F. Rapid Commun. Mass Spectrom. 1995, 9, 453. (17) Montaudo, G.; Montaudo, M. S.; Puglisi, C.; Samperi, F. Rapid Commun. Mass Spectrom. 1995, 9, 1158. (18) Montaudo, G.; Garozzo, D.; Montaudo, M. S.; Puglisi, C.; Samperi, F. Macromolecules 1995, 28, 7983. (19) Montaudo, G.; Scamporrino, E.; Vitalini, D.; Mineo, P. Rapid Commun. Mass Spectrom. 1996, 10, 1551. (20) Garozzo, D.; Impallomeni, G.; Spina, E.; Sturiale, L.; Zanetti, G. Rapid Commun. Mass Spectrom. 1995, 9, 937. (21) Martin, K.; Spickermann, J.; Ra¨der, H. J.; Mu ¨ llen, K. Rapid Commun. Mass Spectrom. 1996, 10, 1471. (22) Spickermann, J.; Martin, K.: Ra¨der, H. J.; Mu ¨ llen, K. Eur. Mass Spectrom. 1996, 2, 161. (23) Dogruel, D.; Nelson, R. W.; Williams, P. Rapid Commun. Mass Spectrom. 1996, 10, 801. (24) Lloyd, P. M.; Scrivener, E.; Maloney, D. R.; Haddleton, D. M.; Derrick, P. J. Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 1996, 37, 847. (25) Deery, M. J.; Jennings, K. R.; Jasieczek, C. B.; Haddleton, D. M.; Jackson, A. T.; Yates, H. T.; Scrivens, J. H. Rapid Commun. Mass Spectrom. 1997, 11, 57. (26) Jackson, A. T.; Yates, H. T.; MacDonald, W. A.; Scrivens, J. H. J. Am. Soc. Mass Spectrom. 1997, 8, 132. (27) Guo, B. C.; Chen, H.; Rashidzadeh, H.; Liu, X. Rapid Commun. Mass Spectrom. 1997, 11, 781. (28) Chen, H.; Guo, B. C. Anal. Chem. 1997, 69, 4399.
Analytical Chemistry, Vol. 70, No. 1, January 1, 1998 131
approach complements conventional techniques. However, it requires a lengthy procedure to collect many fractions and acquires a large number of the mass spectra. Hence, it would be ideal to directly measure the MWD values of an untreated polymer sample from a single MALDI-TOF spectrum, even in the case of polydisperse polymers. The failure of MALDI-TOF in measurements of MWDs of polydisperse polymers is due to mass discrimination against highmass oligomers, and therefore, the measured MWD values are smaller than those data obtained from other conventional methods. Clearly, to overcome the mass discrimination problem, one must understand what experimental parameters of MALDI-TOF could cause this problem and to what extent. Unlike other methods, mass spectrometry is an indirect technique which is based on detection of the gas-phase ions. In general, the MALDI-TOF analysis consists of three major steps. The first involves sample preparation leading to incorporating oligomers of different sizes into the matrix crystals. The second is the MALDI process which converts the polymer analytes into intact, gas-phase ions through laser desorption/ionization. In the third step, a TOF mass spectrometer is used to analyze these formed polymer ions. For MALDI-TOF to be successful in measuring polymer MWDs, the integrated ion peak area or height of each individual oligomer must faithfully represent the molar fraction of that oligomer in the polymer. This requires no mass discrimination against any fractions of the polymer during each of those three processes. Otherwise, MALDI-TOF would fail to provide correct polymer MWDs. As part of a continuing effort to develop the reliable MALDITOF method for polymer analysis, we recently examined the use of MALDI-TOF for measurements of MWDs of polydisperse polymers. The goal of this study is to determine some of the causes of mass discrimination. In this work, we select poly(methyl methacrylates) (PMMAs) for study since the MALDI-TOF analyses of narrowly distributed PMMAs have been intensively studied by our group and many others.
Figure 1. MALDI-TOF mass spectra of a polydisperse PMMA 15K, obtained from the same desorption spot, but using different detection voltages: spectrum a 1500 V; spectrum b 1900 V. The matrix was IAA. Note that the peaks below 1000 Da corresponds to the matrixrelated species.
EXPERIMENTAL SECTION MALDI-TOF Mass Spectrometer. The experiments were performed by using a homemade linear MALDI-TOF instrument. A detailed description of this instrument has been given previously.28 Briefly, a pulsed Nd:YAG laser producing a wavelength of 355 nm was used for MALDI. The desorption laser beam was focused on the sample tip at an incident angle of 45°. The twostage static acceleration voltages were set at 30 and 15 kV, respectively. The used TOF mass spectrometer is ∼1 m long. To reduce the ion signals of matrix-related molecules, most of those ions were deflected away by applying a strong pulsed electric field of 1500 V/cm. One major modification was that a discrete electron multiplier (EMP) (AF850H, EPT Scientific Inc., Auburn, MA) detector was installed to replace the microsphere plate detector (MSP) for this experiment.27 The use of this EMP detector rather than the microchannel plate (MCP) or MSP detectors is based on the consideration that it has a higher dynamic range and a faster gain recovery rate than that of MCP and MSP.29 These features are essential to the MALDI-TOF
analysis of polydisperse polymers. Thereafter, the ion signal was recorded by using a Tektronix 520 digital oscilloscope. All mass spectra were produced in the positive ion mode. A Grams/32 program was used to analyze the recorded data. Materials. The narrowly distributed PMMA 13K (Mn ) 12 500) and PMMA 50K (Mn ) 46 000) are the polymer standards and were provided by American Polymer Standards (Mentor, OH). The polydisperse PMMA 15K (polydispersity 1.5 and Mn ) 8500) was obtained from Aldrich Chemicals (Milwaukee, WI). The narrowly distributed PEG 1450 (Mn ) 1450) and PEG 3350 (Mn ) 3350) were received from Sigma Co. (St. Louis, MO). Two matrixes, 2-(4-hydroxyphenylazo)benzoic acid (HABA)4 and 3-indolacrylic acid (IAA),2 were used in this work. These matrix materials were purchased from Aldrich Chemicals. All the samples were used as received without further purification. Sample Preparation. The polymer solution preparation involved dissolving a polymer sample in acetone or tetrahydrofuran (THF). Unless specified, the concentration of the polymer solution is 1-5 g/L. The matrix solution was prepared by dissolving a matrix material into THF, and its concentration ranges from 5 to 40 g/L. A 1 µL sample of each of the solutions was sequentially placed on a stainless steel probe tip and air-dried before analysis. On several occasions, a small amount of an alkali metal salt such as NaCl was also used to enhance cationization. In those cases, the matrix and salt solutions were first spotted on the tip. After the tip was completely dried, the polymer solution was added to the tip. In this way, the effect of the presence of a second solvent in the polymer solution was minimized.28
(29) Stresau, D.; Hunter, K. Proceedings of the 43th ASMS conference on mass spectrometry and allied topics, Atlanta, GA, May 22-26, 1995.
RESULTS AND DISCUSSION Detector Saturation. Figure 1 shows two spectra of a
132 Analytical Chemistry, Vol. 70, No. 1, January 1, 1998
Figure 2. MALDI-TOF mass spectra of a PEG mixture of PEG 1450 and PEG 3350, obtained from the same desorption spot, but using different detection voltages: spectrum a 1500 V; spectrum b 1900 V. The matrix was HABA.
polydisperse PMMA 15K produced by using IAA as the matrix. Both spectra were obtained from the same desorption spot but by using different detection voltages. Figure 1a was recorded by applying a high voltage of 1500 V to the electron multiplier, while Figure 1b was obtained at 1900 V. The major peaks below 1 kDa correspond to the low-mass matrix-related species. In both spectra, the polymer signals are superimposed onto an unresolved background which is due in part to the matrix-related ions. The presence of a matrix-related background is a common phenomenon when large molecules are analyzed.30 The detailed discussion of this background will be given in the later sections. It is seen that although both spectra were recorded from the same desorption spot, they display two significantly different distributions. Under the low-gain condition (1500 V), the low-mass matrixrelated peaks are relatively weak and the highest PMMA peak is at ∼16 µs in Figure 1a. As the detection voltage was increased to 1900 V, these matrix-related peaks became much stronger. This produced a distinctly different distribution with the highest PMMA peak shifting to 25 µs in Figure 1b. Clearly, this distribution change resulted from detection saturation. For a more quantitative analysis, an experiment was performed by using a mixture of narrow-disperse PEGs 1450 and 3350 as analytes and the HABA matrix. Figure 2 shows two spectra of this mixture, which were obtained from the same desorption spot, but using different detection gains. A voltage of 1500 V was used for Figure 2a, while 1900 V was used to collect Figure 2b. In Figure 2a, the signal of PEG 1450 was much weaker than that of PEG 3350 and the peak intensity ratio of PEG 1450 to PEG 3350 was ∼1:10. As shown in Figure 2b, under the high-gain condition (30) Knochenmuss, R.; Dubois, F.; Dale, M. J.; Zenobi, R. Rapid Commun. Mass Spectrom. 1996, 10, 871.
of 1900 V, the signal intensity of PEG 1450 significantly increased while the ratio of PEG 1450 to PEG 3350 reduced to ∼1:2! It is seen that the strongest peak of PEG 3350 was shifted from 24 µs in Figure 2a to 22 µs in Figure 2b. This suggests that, at the flight time of 22 µs, the ion detection efficiency has been greatly affected by saturation during the course of collecting Figure 2b. This saturation was mainly caused by strong signals of PEG 1450 and low-mass oligomers of PEG 3350. More importantly, the signal of PEG 3350 almost dropped to zero at a flight time of 24 µs in Figure 2b. This indicates that, at that time, the detector was virtually completely saturated! Thereafter, the PEG signal gradually increased in the higher mass range due to the gain recovery. But, this gain recovery was very slow. In Figure 1a, the peak intensity at 17 µs is twice stronger than that at 30 µs, whereas the intensity at 17 µs in Figure 2b was 5-6 times stronger than that at 30 µs. This indicates that, at 30 µs, the detector was far from the complete recovery and that it would take more than 6 µs for this detector to be fully recovered from a major saturation. Clearly, detector saturation would be a problem for polydisperse polymers since with our TOF system, a time window of 6 µs could cover an over 6 kDa mass range at a flight time of ∼80 µs. For an ion detector, there are two different saturation effects. One is the peak saturation due to a high pulse output, and another is the current output saturation due to the limited bias current. In this experiment, we believe that the high current output is the cause of saturation since the peak output is significantly less than 500 mV.29 The ion detector used in this work is one of the better discrete dynode electron multipliers and has a dynamic range more than 10 times better than that of the most advanced MCPs.29 But it still could be underresponse (saturation) to the impacting ions when the average output current reaches to 200 µA. Considering the fact that our detection system has 50 Ω of impedance, it is expected that this EMP could be saturated if the output current is constantly over 10 mV. Polydisperse polymers consist of a large number of oligomers, and the molar fractions of high- and low-mass oligomers are often significantly different. The large difference in molar fraction would not be a problem in GPC since it detects the weight fraction of oligomers. But, this would be a problem in MALDI-TOF since the signal level of high-mass oligomers could be too weak and might be therefore lost in the baseline noise. This, in turn, would result in an underestimation of the fraction of high-mass oligomers. To overcome this difficulty one would have to increase the detector gain. However, to get a good signal to noise ratio in our TOF instrument, the lowest signal level should be ∼1 mV (50 Ω of impedance). This implies that if the molar ratio of lowto high-mass oligomers is 50, the signal of low-mass oligomers should be ∼50 mV to generate a 1 mV signal for high-mass oligomers. Most of the commercial ion detectors have a dynamic range smaller than 200 µA, corresponding to a 10 mV continuous signal output. The polymer signals are quasi-continuous rather than continuous. Hence, we expect that detection saturation would occur when the molar ratio of low- to high-mass oligomers is somewhat larger than 10. It should noted that after we submitted the first version of this paper, we learnt that McEwen and co-workers31 also investigated (31) McEwen, C. M.; Jackson, C.; Larson, B. S. Int. J. Mass Spectrom. Ion Processes 1997, 160, 387.
Analytical Chemistry, Vol. 70, No. 1, January 1, 1998
133
Table 1. Molar Ratio of PMMA 13K to PMMA 50K and the Corresponding Peak Area Ratio matrix
weight ratioa
molar ratiob
area ratioc
errord (%)
HABA IAA IAA
1:10 1:10 1:5
1.7 1.7 3.3
5.6 2.9 5.6
230 70 70
a This is the weight ratio of PMMA 13K to PMMA 50K in the PMMA mixture solution. b This is the molar ratio of the fraction of 8-11 kDa in PMMA13K to the fraction of 48.5-51.5 kDa in PMMA 50K, in the PMMA mixture solution. c This is the average peak area ratio of the fraction of 8-11 kDa in PMMA13K to the fraction of 48.5-51.5 kDa in PMMA 50K, determined from the mass spectra of this PMMA mixture. d Error ) ((molar ratio - area ratio)/area ratio) × 100%.
Figure 3. MALDI-TOF mass spectra of a mixture of PMMA 13K and PMMA 50K obtained by using different matrixes: spectrum a IAA; spectrum b HABA. The weight ratio of this mixture is ∼1:10 of PMMA 13K to PMMA 50K.
instrumental effects in the analysis of polydisperse polymers. They suggested that detection saturation may be responsible for mass discrimination. Clearly, this present work provides strong experimental evidence to support their statement. It appears that even the best ion detectors may not have an adequate dynamic range to avoid this saturation problem. Hence, great efforts must be made to minimize detection saturation. The best solution to this problem is to develop new ion detectors having a high dynamic range. An increase in the detector gain recovery rate will also be helpful in minimizing this saturation effect. Matrix Materials. The matrix plays an essential role in MALDI. Several groups have investigated the matrix effect on analysis of narrowly distributed, smaller PMMAs, and no significant mass discrimination was found in those studies.10,26 In the present work, we examined the effect over a broader, higher mass range. This was motived by the fact that a polydisperse polymer covers a large mass range and that MALDI-TOF presents difficulty in analysis of polymers of over 25 kDa.7 HABA and IAA were evaluated in this work since they are two of the more effective matrixes for narrowly distributed PMMA.4,16 To quantitatively investigate the matrix effect, we used a mixture of PMMAs 13K and 50K to simulate a more polydisperse PMMA. Its distribution covers a mass range from 5 to 70 kDa. Figure 3 shows two spectra of this mixture obtained using IAA and HABA as matrixes, respectively. A weight ratio of 1:10 of PMMA 13K (0.5 g/L) to PMMA 50K (5 g/L) was used to produce those two spectra. It should be pointed out that both spectra were recorded using an optimal detection voltage to avoid saturation. As seen from Figure 3, the use of IAA and HABA as matrix produced two significantly different PMMA distributions and the relative signal ratio of 134 Analytical Chemistry, Vol. 70, No. 1, January 1, 1998
PMMA 50K to 13K was much smaller when HABA was used as matrix. To determine which spectrum would more faithfully represent the distribution of this mixture, an analysis was performed to compare the peak area ratio of these two PMMAs with their molar ratio. In this analysis, we prepared several identical sample tips for each of the used matrixes and three to five spectra were collected from each tip for averaging. This procedure minimized the signal variation effect resulting from the use of different desorption spots or tips, thereby leading to a more accurate determination of the matrix effect. In addition, a new approach was used for this comparison.32 Briefly, this approach involves first determining the moles of a small fraction of a polymer (rather than the moles of the entire polymer) from the MALDI-TOF spectra of this polymer. In this work, the moles of the fractions covering 8-11 kDa (center mass 9.5 kDa) in PMMA 13K, and 48.5-51.5 kDa (center mass 50 kDa) in PMMA 50K were determined, respectively. Thereafter, the peak areas covering these two mass ranges were determined from the MALDI-TOF spectra of the mixture. The detailed description of this procedure will be given in another publication.32 This approach offers several potential advantages. As seen from Figure 3, the left edge of the PMMA 13K peak was overlapped with the matrix-related background. The use of the center portion (8-11 kDa) of this spectrum would reduce this background effect. In addition, each of those two PMMAs covers a large mass range. The selection of a small portion of oligomers from each of PMMAs allows us to examine the matrix effect at two particular masses. Table 1 lists the results arising from this analysis. It is seen that the measured peak area ratio of low- to high-mass PMMAs was larger than the corresponding molar ratio in both cases. This indicates that MALDI-TOF led to mass discrimination against high-mass PMMA oligomers in both cases. The error for IAA was ∼70% while the error for HABA was more than tripled. For IAA, this error could be caused by a combination of the instrumental and matrix effects. The ion impacting energy on the detector was 35 kV (30 plus 5 kV postacceleration). This corresponds to a velocity of ∼1.2 m/s for a 50 kDa ion. This impacting speed was below the threshold of ∼2 km/s.33,34 Hence, it would be reasonable to expect that the ion detection efficiency would decrease to somewhat extent at a mass of 50 kDa. Of (32) Rashidzadeh, H.; Guo. B. C. In preparation. (33) Geno, P. W.; Macfarlane, R. D. Int. J. Mass Spectrom. Ion Processes 1989, 92, 195. (34) Beuhler, R. J. J. Appl. Phys. 1983, 54, 4118.
course, the matrix effect could also contribute to this error. As seen from Table 1, an experiment was conducted to study the concentration effect and we found that an increase in the PMMA 13K concentration by a factor of 2 would not increase or decrease this error. One other conclusion that we could gain from this quantitative analysis is that the contribution from instrumental factors to this error should be less than 70% under the used experimental conditions. This suggests that the use of the matrix HABA was a major source leading to the observed mass discrimination. Like IAA, HABA is one of the more effective matrixes for narrowly distributed PMMAs. For example, Montaudo and co-workers have successfully used this matrix for analysis of PMMA over 90 kDa.14 Hence, this result indicates that even those good matrixes for narrowly distributed polymers may not be suitable for polydisperse polymers. More importantly, this work demonstrates that the use of an improper matrix may be one major cause of mass discrimination encountered in the MALDI-TOF analysis of polydisperse polymers. Therefore, one must be cautious in selection of matrixes for polydisperse polymers. One other problem associated with the matrix is the presence of a strong matrix-related noise background. As shown in Figure 1, when a polydisperse polymer is analyzed, the polymer signals are superimposed on a strong, unresolvable background. In general, two sources could contribute to this background. The first is overlapping polymer peaks due to the poor resolution. The second is caused by the signals arising from matrix-related species. Our MALDI-TOF system can well resolve two neighboring PMMA peaks below 5 kDa. This suggests that the matrixrelated species is one of the major contributors to this background. The presence of a strong matrix-related background was also observed in the spectra of other polydisperse polymers.19 Moreover, it was observed that this background varies across the mass spectra and changes with the oligomer size. This effect was demonstrated in Figure 4. Panels a and b of Figure 4 were obtained in the presence of PMMAs 13K and 50K, respectively. These two spectra were obtained under similar conditions and using the IAA matrix. It is seen that the matrix background increases from Figure 4a to b. Overlapping neighboring oligomer peaks could also produce an unresolved background, especially in the high-mass range. Since it is difficult to distinguish the matrix-related noise from the peak overlapping background, it is virtually impossible to make correction for this noise. Hence, a question is raised concerning to what extent this background affects measurements of MWDs. Currently, a study of the effect of the matrix-related noise is underway. CONCLUSION This work investigated some of the causes of mass discrimination encountered in the MALDI-TOF analysis of polydisperse polymers. It was found that detector saturation could be a major instrumental factor causing mass discrimination. This is due to the strong signals of matrix-related ions and low-mass oligomer ions. Detector saturation could be a fundamental problem since
Figure 4. MALDI-TOF mass spectra of the matrix-related species obtained in the presence of different sizes of PMMAs, respectively: spectrum a 13K; spectrum b 50K. The matrix was IAA.
most of the detectors currently available may not have an adequate dynamic range to avoid saturation, especially when the molar ratio of high- to low-mass oligomers is very small. We also quantitatively analyzed the matrix effect. IAA and HABA were selected for this study. It was found that the use of HABA led to more significant mass discrimination. This shows that the use of an improper matrix may be one other major source of mass discrimination. Hence, a systematic study aiming to searching for better matrixes for polydisperse polymers is needed. In summary, this work demonstrated that both the TOF instrument and the used matrix could cause mass discrimination and that in many cases, those effects can be fundamental limitations. Hence, unless new approaches are developed, one must be cautious in applying MALDI-TOF to directly measuring MWDs of polydisperse polymers, especially those highly polydisperse polymers. ACKNOWLEDGMENT Financial support from Cleveland State University through a RCA grant is gratefully acknowledged. This research is also supported by NIH Grant HG01437. The authors thank the American Polymer Standard Corp. for providing several polymer samples. Received for review June 2, 1997. Accepted October 22, 1997.X AC970568Z X
Abstract published in Advance ACS Abstracts, December 15, 1997.
Analytical Chemistry, Vol. 70, No. 1, January 1, 1998
135
