Tải bản đầy đủ - 0 (trang)
4 Matrix- Assisted Laser Desorption/Ionization (MALDI)

4 Matrix- Assisted Laser Desorption/Ionization (MALDI)

Tải bản đầy đủ - 0trang

FIGURE 9.5

Matrix-assisted laser desorption/ionization FT mass spectrum of PEG 10000. (Reproduced from

ref. 23 with permission of the copyright holder)



trapping plate is set to ground potential and the rear-most trapping plate is

set to its highest value, typically 10 volts, setting up a decelerating field within

the analysis cell. After firing the laser and waiting a fixed amount of time

for ion transit (50 to 300 µs), the electrostatic trap is closed and the polymer

ions remain trapped. This method has become the standard practice for trapping MALDI generated ions in FTMS instruments; MALDI-generated matrix

and analyte ions are ejected with significant velocities and kinetic energies that

scale with mass. The theory of gated trapping has been eloquently described

24

by Knobeler and Wanczek.

25

In some recent work, Easterling et al. have built a 4.7 Tesla internal

MALDI-FTICR instrument. The authors acknowledge that ion introduction

into the homogeneous region of the magnetic field is simplified when ions

are formed in the vicinity of the cell; this eliminates the need for additional

ion lenses, multipole devices, or wire guides. Using other instrument modifications such as an open-ended cylindrical cell with capacitive coupling

and an internal preamplifier, they demonstrate high performance analysis

of singly charged molecular ions in the 1000 to 10000 Da range. Spectra are

shown for PEG 8000 and PEG 4600. The authors propose that, because the

linear flight space between end electrodes is tripled over that of most cubic

©2002 CRC Press LLC



cells, there is less mass discrimination in their case, using an open ended

cylindrical cell over the conventional cubic cell design. The discrimination

issue that affects overall molecular weight distribution calculations will be

discussed further in the mass discrimination section.

The other method for introduction of ions into the magnetic field is the

use of an external MALDI source. Proponents of external ion sources claim

that external ion sources provide the necessary flexibility to interface a variety of techniques such as FAB, SIMS, and ESI. In addition, researchers believe

external sources allow the tuning of instrumental parameters separately,

thereby permitting optimization of the ion formation and detection events

26

separately. An example of this approach was given by Heeren and Boon

who combined “in-source” pyrolysis for broadband screening of polymer

additives with MALDI to characterize molecular weight distribution and

perform end-group analysis. For the determination of an amine terminated

poly(propyleneglycol), the authors also cited extensive discrimination using

the external MALDI ion source. In their experiments, Tgate typically varied

between 600–2000 µs, depending on the molecular weight range of interest.

27

White et al. also developed a new external ion source 7 T FTICR utilizing

an electrostatic ion guide. PEGs of number average molecular weights 1500, 2000,

and 3400 were used as test samples to study time-of-flight effects. By adjusting

the voltages of the ion guide, arrival times could be kept fairly consistent from

1.2 ms for PEG 1500 to 1.5 ms for PEG 3400. By applying quadrupolar axialization (discussed later in this chapter), the authors were able to measure PEG 3400

spanning a several hundred mass unit range with average resolving power of

−9

400,000, due in part to the low (1 × 10 torr) system pressure.

28

Finally, the issue of polymer detection limits was addressed by Pastor et al.

Using PEG 2000 and 6000 standards, they determined a lower detection limit

of 40 femtomoles (from 4 laser shots) and produced a spectrum with a signalto-noise ratio of at least 5:1. Part of the added reproducibility of the MALDI

events came from the use of an aerospray sample deposition technique that

could produce highly uniform sample surfaces. Scanning electron microscopy was used to quantitate the laser spot sizes.

In an effort to prevent duplication of references, all further MALDI references will be included in other areas of this chapter to which they are

appropriate.



9.5



Electrospray Ionization (ESI)



Although not a laser FTMS method, it is necessary to briefly consider the

complementary technique, electrospray ionization, one of the most widely

used alternatives for biological applications where high-mass analysis has

been achieved through the addition of multiple charges. The introduction

of electrospray ionization has revolutionized analysis of high-mass species.

Because of its evaporative ionization process, molecular ion dissociation is

©2002 CRC Press LLC



minimized, thereby yielding multiple charging with correspondingly lower

m/z values. However, very little work has employed the electrospray process

with trapped ion techniques to analyze polymers. An obvious reason is the

lack of control in producing charge states. As the number of trapped ions in

an FTMS is spread over a larger m/z range, the effective sensitivity produced

for any particular charge state is greatly reduced, as a consequence of

decreased relative ion abundances. To make this point more explicitly, consider

29

an example recently published by McLafferty and coworkers. They analyzed a poly(ethylene glycol) 20,000 sample by ESI-FTMS. It was possible to

resolve 5000 peaks from isotopic clusters representing 65 oligomers in 12

charge states. Analysis of data of this magnitude is a daunting task and

clearly would not be the method of choice for polymer analysis, if simpler

alternatives such as MALDI-FTMS are adequate. Additionally, most entrylevel commercial FTMS systems would have a difficult time producing spectra of sufficient resolution to achieve this level of performance.

A more detailed account of ESI-FTMS of polymers was presented by

O’Conner and McLafferty for poly(ethylene glycol)s of 4.5, 14 and 20 kDa.30

Figure 9.6 shows their ESI/FT mass spectrum of PEG 14000 with resolving



FIGURE 9.6

(a) Electrospray ionization FT mass spectrum of PEG 14000 with resolving power of 100,000.

(b) Mr distribution from summed oligomer abundances. (Reproduced from ref. 30 with permission of the copyright holder)

©2002 CRC Press LLC



power of 100,000. This is a highly complex spectrum which includes overlap

of isotopic peaks from different charge states. One benefit of the higher

resolving power of FTMS is identification of polymer impurities. The authors

cite the example of the M 29414+ isotopic cluster. If one of the 294 −CH2CH2O−

groups (44 Da) is replaced with a propylene oxide group −CH2CH(CH3)O−

group (58 Da), then the peak cluster should be centered at 950.6 (extra m/z

+

14/14 ). The abundances of these peaks are <5% of those of M 29414+, demonstrating that the sample contains at most 0.02% (1/294 × 5%) monomer units

containing an extra CH2. As expected, data interpretation is very laborious.

The algorithms employed never seemed to work individually, although combining attributes from several methods seemed to produce acceptable results,

according to the authors.



9.6



Mass Discrimination



The fundamental premise of analyzing polymers using FTMS and the

pulsed nature of a laser is that no mass discrimination exists that could bias

the oligomer intensities and lead to incorrect calculation of molecular

weight distributions. For accurate molecular weight distributions, no biasing effects should be present from the MALDI or cationization process, from

ion transmission to the analysis cell, or from excitation or detection strategies. The first issue of “MALDI discrimination” is a hot topic of debate.

Many people believe that for polydisperse polymers, oligomers at the start

and end of the distribution may be ionized differently either through ion

suppression effects as ions are being ejected into the gas phase or through

preferential cationization. There is evidence that the choice of cation affects

coordination with the polymer. There has been work done by time-of-flight

mass spectrometry analyzing different choices of cations and their effects

upon calculated molecular weight distributions. MALDI analysis of polymers as compared with GPC will continue to be an area of intense interest.

The most obvious form of mass discrimination occurs from ion transmission to the analysis cell and trapping. Because FTMS is a trapped ion technique, it is imperative that the analysis cell accommodate a representative

portion of the polymer being analyzed. It logically follows that ions experiencing a time-of-flight effect may not be properly “sampled” if the trap is

closed prematurely (biasing for lower mass species) or left open too long

(biasing for higher mass species). Even with direct laser desorption studies,

there was concern about accuracy in polymer characterization. Hogan and

31

Laude examined several factors that could lead to possible mass discrimination in LD-FTICR, including laser power density, trapping potential, and

distance between the cell and desorption site. They showed, for example,

that the number average molecular weight for PEG 600, PEG 1000, and PEG

1500 varied by 7, 10, and 12% as the desorption site was displaced over a

10-cm distance from the cell. By varying each of the instrumental parameters,

©2002 CRC Press LLC



they could achieve highly reproducible results that were within a few percent

of the measured GPC values.

Trapping discrimination was also examined for MALDI-generated poly32

mer ions in a cubic cell by Dey et al. The use of a gated decelerating potential

with a single delayed trapping time produces very noticeable discrimination

when examining broad polymer distributions. Their work focused on examining the effect on poly(ethylene glycol) spectra as delay times following

desorption/ionization and before applying static trapping voltages were

systematically varied. Dey et al. proposed post averaging of time-domain

transients to reconstruct the entire broadband polymer distribution. Using a

7 T FTMS-2000 system and a 10 µs sampling interval, the authors reported

a reconstructed spectrum of a synthetic “polydisperse” sample made up of

an equimolar mixture of PEG 1000, PEG 3000, PEG 6000, and PEG 8000.

33

These encouraging results led to further research by Pastor and Wilkins

who concluded that there was a critical window of 2500 Da for the standard

2-inch cubic cell. A polymer that contains oligomers covering a range wider

than 2500 Da will certainly produce discrimination if a single gated trapping

time is used (see Figure 9.7). Polymers within this mass range can be properly



FIGURE 9.7

Spectra of hydroxyl-terminated polybutadiene 1350, taken at different gated trapping deceleration times following the laser pulse. Longer delays before applying static trapping voltages

clearly show the time-of-flight effect of the ions entering the cell. (Reproduced from ref. 33 with

permission of the copyright holder)

©2002 CRC Press LLC



represented by optimizing the polymer’s ion abundance over a small incremental time range. Thus, a single optimized trap time can be used to analyze

polymers of narrow polydispersity. Employing a 3 T modified FTMS-1000

mass spectrometer equipped with a nitrogen laser, Pastor and Wilkins demonstrated characterization of both narrow and wide mass nonpolar polymers

up to m/z 6000, including several polyisoprene, polybutadiene, and polystyrene samples.

Other groups have noted that the time-of-flight effect of the ions is a definite

34

problem particularly when an external ion source is used. Easterling et al.

reported a reduced sensitivity to this discrimination effect by use of internal

MALDI and an open-ended cylindrical analyzer cell of over three times the

length of a standard cubic cell. They reported a mass shift of only 20 Da for

a PEG 1000 polymer distribution over a 200 µs change in gated trapping

time. Computer simulations showed a much improved field free region

owing to the extended length of the flight path. Another method of constructing the “true” polymer distribution from its time segmented parts was

35

proposed by O’Conner et al. who advocate superimposing spectra on the

same m/z axis rather than summing data transients. Their method is based

on “acquiring a series of spectra at different trapping times and superimposing the spectra so that each oligomer has the intensity of its maximum

intensity through the set of spectra.” Use of a chromatographic polystyrene

standard (MW 950) produced an error of 20% for the most probable polymer

weight, Mp.



9.7



Quadrupolar Excitation/Ion Cooling



One of the latest innovations in FTMS has been the introduction of quadrupolar

excitation/ion cooling (also known as quadrupolar axialization or ion axialization). Introduced as an ion manipulation technique, axialization can be used to

compress ion clouds within the analysis cell, allowing for a tighter, more coherent motion during the excitation/detection events. A secondary benefit of

applying axialization is its effect on nonselected ions, which are lost under high

pressure through magnetron expansion of their orbits. The technique works by

applying low sinusoidal in-phase voltages (typically 5 volts or less) to opposite

pairs of cell plates. During the quadrupolar excitation event, magnetron motion

is coupled to the cyclotron motion of the ions. In the presence of a collision gas,

ions are cooled to the center of the cell both through cyclotron and axial relaxation. Magnetron relaxation, for ions not undergoing the applied frequency,

causes a radial expansion of the remaining ions’ orbits and eventual loss from

the cell. This process can be used to accomplish highly specific mass selection.

To date, axialization is an accepted technique for improving signal-to-noise

ratios, mass resolving power, ion capture efficiency, and ion remeasurement

36

efficiency. The technique has been reviewed by Guan et al.

©2002 CRC Press LLC



37



Showing interest in manipulating polymers, Pastor et al. used quadrupolar excitation in a 7 T FTMS-2000 equipped with a dual cubic cell. First,

the source cell was used to apply the axialization to the oligomers of

interest. Following this, ions were transferred to the analyzer cell for detection. Using PEG 6000 as a test polymer, the authors demonstrated that the

mass selection could be very precisely controlled by varying the applied

amplitude of the single frequency quadrupolar excitation. Figure 9.8 shows



FIGURE 9.8

(a) MALDI FT mass spectra for PEG 6000 showing full distribution using source cell detection.

(b) Mass spectrum of a selected oligomer using quadrupolar axialization to isolate and transfer

to the analyzer cell. (Reproduced from ref. 37 with permission of the copyright holder)



©2002 CRC Press LLC



the selection of a single oligomer from the polymer PEG 6000. This demonstrated an extremely effective way of reducing ion populations. It is a

well-known fact that examining an entire polymer distribution can lead to

reduced resolving powers at high trapping voltages, used to improve

signal-to-noise, due to space charge interactions. A common solution is to

reduce the overall number of trapped ions, leading to longer-lived transients with fewer interactions from other ions and, ultimately, better resolving power. In the past, to arrive at the same mass selection, chirp ejection

pulses were applied to the polymer or the ions were subjected to stored

38,39

waveform inverse Fourier transform (SWIFT).

However, both frequency

sweep and SWIFT ion ejection techniques affect the radius of the analyte

ion(s) of interest, and they usually require several applications of tailored

excitation to cleanly select the peak(s) of interest. On the other hand,

axialization performs the selection in a single step. Pastor et al. established that

axialization could be applied to polymers with masses as high as 13,000 Da,

the highest mass tested.

40

In a further development of axialization, Marto et al. demonstrated broadband axialization of polymers using an external MALDI source and an electrostatic ion guide (EIG) in a 3 T FTICR. Broadband axialization can improve ion

trapping efficiency of injected ions from an external source. The benefit is

reduced off-axis ion displacement. It can also be used to select a subset of the

polymer distribution, also achievable with single frequency axialization, but

without the oligomer abundance distortions. Thus, the spectrum over the

selected mass range should closely match the relative abundances of the original distribution. To achieve the broadband effect, Marto et al. used repeated

SWIFT excitations controlled from a Macintosh II personal computer and a

short C language program. For PEG 2000, an axialization range of 1950 to 2120

Da was selected, corresponding to selection of 5 oligomeric species.

41

Pitsenberger et al. examined alternative excitation mechanisms for ion axialization and remeasurement in a 4.7 T FTMS using internal MALDI. Specifically, they looked at repetitive chirp, filtered noise, and high-amplitude single

frequency excitation for ion axialization. Because the instrument employs an

internally mounted preamplifier on the detection electrodes, said to improve

overall S/N and “reduce the adverse effects of distributed capacitance on the

image current,” the authors instead applied two-plate axialization, a modified version of the axialization technique. For frequency chirp broadband

axialization, a mass-to-charge range of 1500 to 2500 was selected for PEG

2000 with remeasurement efficiencies exceeding 99.5% using the open-ended

cylindrical cell. The filtered white noise experiments yielded similar efficiencies

but for a smaller mass range, 400 Da, due to inadequacies in the active bandpass filter used, which peaked at m/z 2000. Finally, the high-amplitude single

frequency experiments, when combined with capacitive coupling of the excitation signal to the trapping plates, produced less axial ejection of ions and a

wider retained range of ions. However, the distributions of axialization-selected

oligomers still suffer from the uneven power applied during axialization,



©2002 CRC Press LLC



leading to distributions that center around the applied frequency. These

authors also obtained spectra of PEG 4600, PEG 6000, and PEG 8000 that were

collected using frequency burst excitation for ion axialization. These spectra

appear remarkably similar to the PEG 4600 and PEG 8000 shown in the prior

25

Easterling et al. reference that were said to result from “single laser shots

followed by the amplified trapping….”

Justification for using two-plate quadrupolar excitation has been reported

42

by several groups including Jackson et al. who demonstrated the qualitatively similar results of a mixture of PEG 1000 and PEG 1500 undergoing

both two-plate and four-plate ion axialization. The authors note that despite

the complex dynamics and interacting resonances produced from a twoplate excitation, the results show similar efficiencies in their ability to cool

ions.

The effect of capacitive coupling on quadrupolar excitation (but using two

43

plate axialization) was further demonstrated by Pitsenberger et al. These

authors were able to remeasure the 41-mer of PEG for up to 200 remeasurement cycles with 100% efficiency in a capacitively coupled cylindrical ion

trap. Plots were shown demonstrating that the predominant loss of ions occurs

through axial ejection during excitation. Lowering the excitation radius to

prevent radial ejection, along with raising the trapping voltages to prevent

axial ejection of low masses, makes the 100% remeasurement possible. Broadband remeasurement of a 600 m/z range for PEG 2000 produced a 99.5%

efficiency after 50 remeasurements.

Finally, the ability to manipulate polymer distributions and reduce ion

populations was further demonstrated by low-voltage on-resonance ion

44

selection (LOIS). This newly introduced technique uses the same high pressure collision gas as quadrupolar axialization but does not require the hardware switching of the latter. Thus, it is an easily implemented technique on most

Fourier transform mass spectrometers and can offer some of the same advantages as ion axialization. Pastor and Wilkins showed high selectivity of several

poly(ethylene glycol) samples ranging in mass from 1000 to 6000 Da. In some

examples, oligomers were selected in various combinations, including alternating oligomers, and groups to demonstrate that any combination of ions

could be mass selected. LOIS has also shown promise in ion remeasurement

45

studies. The theory is currently being investigated.



9.8



Direct Applications



Applications of specialty polymers are numerous. Of particular analytical interest are copolymers. Here, recent literature on laser desorption

FTMS analysis of non-PEGs, PEOs, PSs, and PMMAs will be discussed

briefly. These articles, which have appeared over the last 10 years, provide

©2002 CRC Press LLC



examples of the variety of polymer systems that can be analyzed by

FTMS.

46

Srzic et al. studied natural polymers of humic acid and lignins by LDFTMS. Humic substances are commonly found in soil and water, and formed

by the chemical and biological degradation of plants and animals. The resulting products can associate into complex organic structures. Lignin is a major

cell wall component in wood and is composed of substituted phenylpropane

units connected through ether links. Spectral peaks up to m/z 700 were

observed for positive ions of humic substances collected from lake sediments. Samples of lignin from birch and spruce were more interesting

because they showed the 444 Da repeat of a trimer building block and masses

up to 3000 Da in negative ion mode.

LD-FTICR-MS was compared with high performance liquid chromatography (HPLC) for analysis of triton polymers, used as commercial surfactants.

47

Liang et al. found that octylphenol ethoxylates gave molecular weight distributions up to 3500 Da, with minimal fragmentation. The Triton X family

of surfactants are commonly used in industrial cleaners and detergents, manufacturing processes, and agricultural applications. The authors show direct

comparison with HPLC/UV chromatograms and LC/quad mass spectra

(Figure 9.9). Overall, LD-FTICR-MS overcomes problems with fragmentation

of higher molecular weight species, reduces analysis time, and provides fully

resolved oligomer information.

An interesting method was presented for the analysis of perfluorinated

48

polyethers (PFPE) by Cromwell et al. The fluorinated polymers are commonly used as lubricants because of their low vapor pressures, chemical

inertness, and thermal stability. Samples were examined in a cubic trap

FTMS-2000 system using a 4.3 T magnetic field. Both a Kr/F excimer laser

and a Nd/YAG laser were used for internal LD. The two-step procedure

involved using the 248-nm excimer laser for desorption of the polymer and,

immediately following, a second more tightly focused 532 nm laser ablation

pulse to ablate metal cations from the surface beneath the polymer sample.

Thus, cationization could be controlled, and the dynamics of the process are

discussed. Mass spectra extended up to 10,000 Da are shown.

A variety of electrochemical polymerization studies have appeared. Elliott

49

et al. used LD-FTMS for the analysis of 4-methyl-4’-vinyl-2,2’-bipyridinecontaining metal complexes. Mass spectral analysis in conjunction with thin

layer chromatography (TLC) demonstrated that normal “polyvinyl-type”

50

chains are formed through chain propagation. O’Malley et al. used LD-FTMS

to study chemically and electrochemically prepared poly(2-vinylthiophene).

They identified three forms of poly(2-vinylthiophene): (i) free radical-initiated,

(ii) a form arising from anodic oxidation of 2-vinylthiophene, and (iii) an insoluble film on the anode. Using high resolution FTMS, these workers drew

conclusions about different mechanisms involved in the formation of each of

the polymers.

51

Campana and co-workers wrote a general tutorial on polymer analysis

providing an excellent introduction to LD-FTMS.

©2002 CRC Press LLC



FIGURE 9.9

(a) LC/UV chromatogram of Triton sample, (b) LC/quadrupole mass spectral data of Triton

sample, (c) LD-FTICR mass spectrum of Triton sample taken from a single time-domain data

set. (Reproduced from ref. 47 with permission of the copyright holder)



9.9



Polymer Additives



It would be remiss not to discuss polymer additive analysis. Such materials

are used as antioxidants, UV absorbers, and stabilizers, and dyes. An early

paper dealt with LD-FTMS analysis of poly(methyl methacrylate) dyes such

as 12-H-phthaloperine-12-one (orange dye), 1-(methylamino)anthraquinone

52

(red dye), and 1,8-bis(phenylthio)anthraquinone. Hsu and Marshall

©2002 CRC Press LLC



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

4 Matrix- Assisted Laser Desorption/Ionization (MALDI)

Tải bản đầy đủ ngay(0 tr)

×