Polymer synthesis often results in impure products, containing
multiple end groups, due to the synthesis process itself or the presence
of impurities. Identification and quantification of these end groups
and impurities is important because they influence polymer properties.
An interlaboratory comparison of a mixture of two polystyrenes, one
terminated with a proton, and the other terminated with a hydroxyl
group, was performed using matrix-assisted laser desorption-ionization
mass spectrometry (MALDI-MS). This interlaboratory comparison will
be used to quantify the accuracy and repeatability of MALDI-MS in
the identification of multiple end groups as a result of polymer synthesis.
|
|
Charles M. Guttman and
William E. Wallace |
|
|
A MALDI- MS interlaboratory comparison of mixtures of synthetic
polymers of the same repeat unit and with closely matching molecular
masses but with different end groups was sponsored by NIST. This interlaboratory
comparison was designed to determine how well a group of laboratories
engaged in MALDI-MS analyses of synthetic polymers agree on the percent
composition of each polymer in the mixture. In an earlier interlaboratory
comparison, NIST studied the molecular mass distribution (MMD) and
the moments obtained by MALDI on polystyrene (PS) and compared the
results with those obtained by classical methods. Interlaboratory
comparisons enable us to gauge the needs of the community for standard
methods and materials, to begin a dialog with the community on these
needs, and to steer our own research. The interlaboratory comparison
on mixtures of two polystyrenes with different end groups is of particular
interest to industry where understanding and controlling the chemistry
of reactive pre-polymers is a major concern. |
|
Five gravimetric mixtures of two polystyrenes with matched molecular
masses but with different end groups, -H and -OH, were created. The
challenge for the participants was, without knowing the gravimetric
values, to determine the mass fraction of each polymer in each mixture.
Difficulties to overcome include different ionization probabilities
of the two polymers, and finding small peaks in background noise for
species that are in the minority in the mixture. Fourteen labs from
5 countries participated: 6 industrial, 4 academic, and 4 government
laboratories. |
|
The PS syntheses were both initiated with butyl initiator. One was
terminated with H (PSH) and the other was terminated with CH2CH2OH
(PSOH). These end groups are separated by 44 u and 60 u in the mass
spectrum and are well resolved (Fig. 1). Five mixtures of these were
made up with mass ratios between approximately 95:5 to 10:90 PSOH:
PSH. Nuclear magnetic resonance (NMR) and Fourier-Transform infrared
(FTIR) spectroscopies conducted at NIST confirmed the gravimetric
mass ratio. The existence of a small amount of PSH in the as-received
PSOH material was evident in our MALDI-MS analysis. NMR gave between
(4 to 9) % PSH polymer in the PSOH-material. This is close to the
value obtained from the initial MALDI-MS analysis performed at NIST.
|
|
|
Figure 1: MALDI mass spectrum of polystyrene mixture. |
|
Each participating laboratory was asked to perform MALDI-MS on each
of the five mixtures using two distinct protocols: one using all-trans
retinoic acid with defined concentrations of matrix:analyte:salt and
a second protocol defined by the user as their preferred method for
analyzing polystyrene. Each laboratory was asked to report two repeats
of each protocol for each mixture to check for intralaboratory variability.
Each laboratory was asked to report the estimated ratio of the mass
of PSH to the mass of PSOH in each mixture and return data sets listing
mass versus the integrated peak area of the mass spectra. |
|
The analysis of the data revealed a strong variation among laboratories.
This was seen in our first interlaboratory comparison as well. The
effect of instrument manufacturer on the data was not significant.
But the difference between acquiring mass spectra in the linear and
the reflectron mode of operation resulted in significant variation
in the end group fractions. The analysis of variance showed no significant
influence of the sample preparation protocol on the fraction of PSOH
in each mixture. |
|
The accuracy of the instrument calibration of each laboratory was
assessed by calculation of end group mass. In our first interlaboratory
comparison, calibration varied more widely than anticipated. We would
expect the calibration of most time-of-flight mass spectrometers in
this mass range to be off by less than 3 u. Some laboratories showed
calibrations off by over 40 u. Figure 2 shows that calibration was
again an important issue in this interlaboratory comparison. The instrument
calibration varied by as much as 20 u, even though the end groups
were known |
|
|
Figure 2: The average molecular mass of PSH end groups
for each laboratory. The red line represents the actual molecular
mass of the end groups. |
|
Figure 3 shows a plot of the "Theoretical Mixture of PSOH"
polymer fraction assuming 7 % PSH polymer in the PSOH material from
the NMR data. The plot is compared to MALDI-MS data from some of the
participating laboratories. The agreement is generally good, but the
data show a wide range of variability among the laboratories. Deviation
of the data from the expected percentages was found widest at the
extremes of concentration of 95:5 and 10:90 PSOH: PSH, but even at
the 50:50 ratio the deviation was as high as 10 %. |
|
|
Figure 3: The fraction of PSOH determined by MALDI compared
with the gravimetric data. |
|
The agreement and standard deviation of data are shown in Table
1. The fraction of PSOH polymer in each mixture is given in the 2nd
column. This column assumes the PSOH-material is 7 % PSH polymer and
93 % PSOH polymer. Seven percent is the mid point of the range assigned
by the NMR. It is seen from both Fig. 3 and Table 1 that the agreement
between the MALDI-MS and the gravimetric results for the fraction
of PSOH polymer in each mixture is good when judged by the standard
deviation of the MALDI results. |
|
|
By Mass
|
MALDI
|
MALDI
|
|
|
Average
|
Std Dev
|
Mix A
|
0.93
|
0.933
|
0.051
|
Mix B
|
0.716
|
0.746
|
0.086
|
Mix C
|
0.481
|
0.47
|
0.041
|
Mix D
|
0.284
|
0.259
|
0.051
|
Mix E
|
0.086
|
0.117
|
0.2
|
|
Table 1: The fraction of PSOH polymer by mass and determined
by MALDI. |
|
The key to "getting it right" in MALDI-MS, especially
for Mix A and Mix E, is to optimize the instrument. This was seen
in the higher variation in the fraction of PSOH determined for Mix
A and Mix E, where a minor component may be lost in the noise. The
result, which indicates significant variation in the fraction of PSOH
for instrument mode, may be due to the resolution difference between
reflectron and linear mode. In order to better understand the optimization
of the MALDI-MS we have begun to study the influences of the instrument
and sample preparation parameters on the MALDI-MS signal. We are using
an orthogonal fractional factorial design to study detector voltage,
laser energy, delay time, and ion optic voltages, as well as matrix
and polymer concentration. Once the effects of all parameters are
understood, optimization will be easier to obtain. |
|
These results also recapitulate the need for an unbiased, automated
peak integration routine when the S/N is poor. We will continue to
refine our MassSpectator software as a benchmark method to perform
unbiased peak integration. |
|
|
For More Information on this Topic |
|
S.J. Wetzel, K.M. Flynn, B.M. Fanconi, D.L. VanderHart (Polymers
Division, NIST); S. Leigh (Statistical Engineering Division, NIST)
|
|
C.M. Guttman, S.J. Wetzel, K.M. Flynn, B.J. Fanconi, W.E. Wallace,
and D.L. VanderHart, "International Interlaboratory Comparison
of Mixtures of Polystyrenes with Different End Groups ," in Proceedings
of the 52nd ASMS Conference on Mass Spectrometry and Allied Topics
, Nashville, TN. |
|
C.M.Guttman, S.J. Wetzel, W.R. Blair, B.M. Fanconi, J.E. Girard,
R.J. Goldschmidt, W.E. Wallace, D.L. Vanderhart, Analytical Chemistry
73(6), 1252-1262 (2001). |
|
S.J. Wetzel, C.M. Guttman, K.M. Flynn, "The Influence of Matrix
and Laser Power on the Molecular Mass Distribution of Synthetic Polymers
obtained by MALDI", Int. J. Mass Spectrometry, in press. |
|
|
|
|
|
|
|
|
|