Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
MALDI/TOFMS method for general purposes because
high performance systems having high mass resolu-
tion are required for analyzing complex and detailed
mass spectra.
Although few reports exist on the SEC/ESIMS
method utilizing quadrupole mass spectrometers (q-
MS) having low mass resolution 7), important informa-
tion such as molecular weight and constitutional
repeating units, which are required for structural analy-
ses of synthetic polymers, should be obtained if limit-
ing the analysis targets to components having relative-
ly low molecular weights (up to around 10,000). In this
paper, we will explain what level of analysis can be
achieved using a q-MS by introducing spectrum analy-
sis methods and measurement procedures.
Furthermore, this paper consists of revised versions
of the reports released at the 54th Annual Conference
on Mass Spectrometry Osaka Japan 2006 and at a
workshop for mass spectrometry entitled “Recent
development of synthetic polymer analysis” 8) and the
9th and 11th National Symposium on Polymer Analysis
and Characterization 9), 10). The progress of systems
and methods in the field of synthetic polymer mass
spectrometry is extremely rapid. Please refer to the lat-
est data and documents for the current situation.
Information on SEC/ESIMS
SEC (size exclusion chromatography) is a liquid
chromatography technique which separates sample
substances according to their molecular sizes. It elutes
Introduction
For mass spectrometry of synthetic polymers,
MALDI/TOFMS1) is commonly used for general pur-
poses due to its simplicity. Although a matrix is used as
an ionization aid in the MALDI method, some trial and
error is necessary to apply the method to the analysis
of synthetic polymers and synthetic polymer com-
pounds having unknown structures, as an appropriate
matrix suitable for the target chemical substance(s)
must be carefully selected. Furthermore, samples hav-
ing large molecular weight distributions (Mw/Mn) are
not readily ionized, and such samples require pretreat-
ment refining. Therefore, in the MALDI method, it is
necessary as a principle to combine with a separate
fractionation system because online combination with
chromatography is difficult.2)
On the other hand, the ESI method 3), 4) is an effec-
tive ionization method for chemical substances having
high ionicity or polarity. Through this method,
hydrophobic synthetic polymers can also be ionized by
using an alkali metal salt or other similar substances as
a cationization agent, thus enabling a mass spectrum to
be obtained.5) Moreover, online combination with chro-
matography can be readily conducted in this method
because ionization is performed by spraying a sample
solution. This ionization method is most commonly
used in the LC/MS method. Therefore, some structur-
al analyses using SEC/ESIMS were reported in the
early 1990s.6) However, at the present time, the
SEC/ESIMS method is not as commonly used as the
1SUMITOMO KAGAKU 2008-II
Investigation of Synthetic PolymerCharacterization by means ofSEC/ESIMS
Sumitomo Chemical Co., Ltd.
IT-Related Chemicals Research Laboratory
Yoshinobu TSUCHIDA
Keiko YAMAMOTO
Hiromi YAMADA
It is difficult to analyze synthetic polymers using mass spectrometry, because they have a very wide moleculardistribution (Mw/Mn). But if we select a relatively low molecular species, we can obtain information on polymerstructure by means of MS. In this paper we describe an analysis method for synthetic polymers by means ofSEC/ESIMS (using q-MS), some analysis cases, and the problems of this method.
This paper is translated from R&D Report, “SUMITOMO KAGAKU”, vol. 2008-II.
substances having larger molecular sizes faster than
those with smaller sizes. It is more commonly referred
to as GPC (gel permeation chromatography). However,
in this paper, we will refer to this method as SEC
because this expresses its separation principle.
In the SEC/ESIMS method, SEC and ESIMS are
combined online 6), 7). Fig. 1 shows a block diagram of
the system used for the investigation that we conduct-
ed in this case. For the cationization agent, the post col-
umn addition method --in which the agent is added
after the column is eluted-- was used.
Evaluating Cationization Agents
In the SEC analysis of hydrophobic synthetic poly-
mers, THF (tetrahydrofuran) and chloroform are com-
monly used for the mobile phase. For the cationization
agent, NaI and KI are commonly used due to their solu-
bility in the mobile phase.5) In order to obtain a cation-
ization agent that can also substantially ionize macro-
molecular components, we compared KCl, NaI and KI.
The cationization agent was a 1:1 mixed solvent of
water and acetonitrile at a concentration of 1mM. The
mixed solvent was then added at a rate of 0.05mL/min
using the post column addition method. The samples
used for the evaluation were mixtures of different mol-
ecular weights of PMAA (poly-methylmethacrylate,
Fig. 2) from Polymer Laboratories, Inc., which was
commercially available as a molecular weight calibra-
tion reference standard,The molecular weights (MW)
used for mixing were 3100, 6540, 9400, 12700 and
29300. The concentration was 1mg/mL and the injec-
tion rate was 25µL.
Fig. 3 shows the chromatogram and Fig. 4 shows
the mass spectrum in the area A (with no background
processing).
2SUMITOMO KAGAKU 2008-II
Investigation of Synthetic Polymer Characterization by means of SEC/ESIMS
Fig. 1 Schematic Diagram of SEC/ESIMS Sys-tem
SEC System(Agilent 1100 LC)Agilent 1100 LC/MDS
0.25mL/min THF
50µL/min
LC Pump for addition
Fig. 2 PMMA (Poly-methylmethacrylate)
C5H8O2••
Exact Mass : 100.05243Mol. Wt. : 100.11582
C10H18O4••
Exact Mass : 202.12051Mol. Wt. : 202.24752O
OO
OO
O
n
Fig. 3 Size-exclusion Chromatograms of PMMA mixture
a)
b)
c)
d)
a) 220nm. b) TIC, +NaI. c) TIC, +KI. d) TIC, +KClA ; averaging area of spectrums in Fig. 4Column ; Tsk-gel Super HZ(4000*3+3000+2500*2)
A
mAU
02468
10
010000002000000300000040000005000000
0100000020000003000000400000050000006000000
100000020000003000000400000050000006000000
min45 50 55 60 65 70
min45 50 55 60 65 70
min45 50 55 60 65 70
min45 50 55 60 65 70
Fig. 4 SEC/ESIMS Spectrum at A
020406080
100
020406080
100
m/z500 1000 1500 2000 2500
m/z500 1000 1500 2000 2500
m/z500 1000 1500 2000 2500
020406080
100 a)
b)
c)
a) +NaI. b) +KI. c) +KCl
�
�
�
��
� �� � � �
�
��
�
�
�
� � �
�
� ��
�
472.
7
622.
6
334.
9
922.
3
772.
4
1072
.2
413.
3
1222
.1
1971
.6
1521
.9
1372
.0
1671
.8
545.
3
1821
.7
301.
1
2342
.3
934.
4
2068
.8
830.
4
2225
.4
634.
7
2509
.5
2721
.226
29.2
2906
.228
16.2
408.
3
726.
6
Max : 2538
536.
6
370.
7
702.
4
1034
.2
2041
.6
2141
.8
1941
.5
564.
6
1791
.4
429.
2
704.
4
2242
.2
2342
.4
1531
.7
1200
.0
1365
.9
315.
1
868.
3
2442
.4
1704
.0
2575
.9
1062
.2
2862
.7
2052
.321
40.3
2722
.1
566.
6
1890
.0
459.
2
2982
.4
732.
7
1799
.7
2141
.7
2041
.7
1941
.5
2342
.0
2242
.3
2113
.3
2442
.0
1841
.5
2013
.1
2213
.8
2313
.5
336.
7
2542
.3
1006
.1
2642
.0
1741
.5
2104
.3
1912
.7
2413
.6
710.
3
2742
.6
429.
2
560.
5
2194
.0
2304
.2
2003
.0
2862
.3
313.
1
1641
.0
1229
.9
858.
2
2982
.6
1531
.6
1381
.6
1128
.2
980.
2
2572
.7
1810
.2
Max : 1986
Max : 2940
As a result, although there was no large difference
observed in TIC (total ion chromatogram), substantial
differences were observed in the mass spectra. For
NaI, a group of cluster peaks (�) caused by NaI and
with intervals of 150Da was detected more strongly
than the PMMA peak(indicated with a solid red line),
which was considered to be the background spectrum.
For KI, a group of cluster peaks (�) caused by KI and
with intervals of 166Da was detected more strongly
than the PMMA peak.
Conversely, for KCl, the background peaks were
weak (��) and the PMMA spectrum was detected as
the primary peak group. It was clear that the anionic
part of the cationization agent greatly affected the back-
ground spectrum, thus influencing the ionization of the
analysis target.
On the metal surface of the ESI probe tip, to which a
high voltage was applied, it can be assumed that the
following reaction with anions occurs: since Cl2 has
higher volatility than I2, it can be more readily dis-
charged outside the system (outside the mobile
phase), thus producing only a minimal amount of clus-
ter ions. Therefore, it is assumed that the sample itself
can be more readily ionized.
2Cl– – 2e– ➝ Cl2 2I– – 2e– ➝ I2
As a result of our investigation --we will not describe
the details here--, it has been clarified that both KCl
and KOH show the highest sensitivity to high-molecu-
lar-weight components and produce simpler spectra.
SEC/ESIMS Spectra of Synthetic Polymers
In general, when measuring high-molecular-weight
components using the ESI/MS method, an MS spec-
trum with multiply-charged ions can be obtained.11) In
the SEC/ESIMS spectra of hydrophobic synthetic poly-
mers when using a cationization agent, multiply-
charged ions to which several cationization agents are
added can be detected. In order to estimate the molec-
ular weight of the target substance based on the
SEC/ESIMS spectra, it is necessary to analyze the
peaks of these multiply-charged ions.
When considering the PMMA molecule as M, and
measuring it in a positive ion mode using KCl as the
cationization agent, the PMMA becomes ionized in the
form [M+zK]z+. z indicates the charge number. In this
case, it is equivalent to the number of K+ ions added
when the molecular weight is rather small, z=1. Howev-
er, when the molecular weight is large, it is detected as
multiply-charged ions in which z ≥ 2.
The threshold at which multiply-charged ions appear
differs depending on the type of chemical and cation-
ization agent. In the case of PMMA, divalent ions begin
to appear once n=20 and the molecular weight is
approximately 2000 (Fig. 2). As the molecular weight
increases, the number of charge numbers detected
increases to trivalent and further to tetravalent. In addi-
tion, these can be detected in the form of a sequence
of peaks, such as divalent, trivalent and tetravalent.
Figs. 5 and 6 show the actual measurement results.
As shown in Fig. 6, an ESIMS spectrum that does
not use SEC appears to be a single sheet diagram in
which all the spectra shown in the graphs B1–B4 are
overlaid. For this reason, it is generally considered that
multiply-charged ion spectra are complex and hard to
analyze.
However, when using SEC, the samples are mea-
sured in order of their molecular weights, thus enabling
the overlaid multiply-charged ions to become separated
from each other. As a result, the spectrum becomes
simpler, thus leading to easier analysis. However, as
shown in B1, it is still difficult to conduct SEC analysis
of high molecular weight areas having low separation
capability. On the other hand, this can also be evidence
that PMMA is ionized. Therefore, it can be considered
that once the separation capability of SEC is improved,
molecular weight estimation of high-molecular-weight
components will be possible.
The x-axis of a mass spectrum is m/z (mass-to-
charge ratio). m indicates the molecular weight of the
detected ion, and z indicates the charge number
3SUMITOMO KAGAKU 2008-II
Investigation of Synthetic Polymer Characterization by means of SEC/ESIMS
Fig. 5 Size-exclusion Chromatograms of PMMA mixture
a) 220nm b) TIC, 1mM KCl +0.05mL/min B1~B4 ; averaging area of each mass spectrum in Fig.6B1 ; 49.0 – 49.2min B2 ; 51.5–52.0min B3 ; 57.0 –58.0min B4 ; 61.0 –62.0Column ; Tsk-gel Super HZ(4000*3+3000+2500*2)
a)
b) B1B2
B3
B4
mAU
05
1015
100000020000003000000400000050000006000000
min40 45 50 55 60 65 70 75
min40 45 50 55 60 65 70 75
(=charge state) as described previously.
In Fig. 6, we examined a PMMA molecule with n=55
(molecular weight 5708.6). The molecular weights of
the K+-added divalent and trivalent ions are shown
below. The molecular weight was used for the calcula-
tion, not the precise mass:
5708.6 + 2 × 39.1 = 5786.8 ------ [M+2K]+
5708.6 + 3 × 39.1 = 5825.9 ------ [M+3K]+
However, as the mass spectrum is detected using
the mass-to-charge ratio (m/z), peaks are detected at
the positions of 2893.4 [M+2K]+ and 1942.0 [M+3K]+.
The measured values shown in Fig. 6 are quite consis-
tent with the values obtained by the calculation.
Thus, in order to estimate the molecular weight from
the multiply-charged ion spectra (peaks), the charge
number must be determined. However, when measur-
ing macromolecules having completely unknown struc-
tures, the charge number must be determined from the
spectrum only.
Determining Charge Number Using Isotope
Peaks
Generally, charge numbers are determined by the
peak intervals between isotopes. Fig. 7 shows enlarged
[M+K]+ and [M+2K]2+ peaks of PMMA (n=27). The
mass spectrum peaks indicate the distribution of stable
isotopes corresponding to the numbers and types of
elements contained in the detected chemical com-
pound. As shown in a) of Fig. 7, several peaks appear,
showing a pattern which reflects the isotope distribu-
tion (abundance ratio). When measuring a 1+ charged
ion (i.e., z=1), the interval of these peaks (m/z interval)
is inversely proportional to the atomic mass unit (1Da).
Therefore, for a 2+ charged ion (z=2), the atomic mass
unit is 0.5Da; for a 3+ charged ion, the unit is
1/3Da=0.333…; and for a tetravalent ion 4+ charged
ion, the unit is 0.25Da. However, as shown in b) of Fig.
7, because a spectrum measured using a q-MS having
low mass resolution does not possess adequate mass
resolution capability, even the 2+ charged ions cannot
4SUMITOMO KAGAKU 2008-II
Investigation of Synthetic Polymer Characterization by means of SEC/ESIMS
Fig. 6 SEC/ESIMS (+KCl) Spectrums at B1 to B4
1000 1250 1500 1750 2000 2250 2500 2750 3000 m/z
1000 1250 1500 1750 2000 2250 2500 2750 3000 m/z
1000 1250 1500 1750 2000 2250 2500 2750 3000 m/z
1000 1250 1500 1750 2000 2250 2500 2750 3000 m/z
n=24n=25
n=26
n=27*
n=23
n=22
n=21
n=26
n=27*
n=28
n=25
n=24n=23
n=22
n=21
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
1808
.4
1708
.1
1875
.0
1641
.5
2492
.925
42.5
2442
.5
2593
.0
2392
.5
1941
.8
2643
.0
2342
.4
2693
.127
42.9
1574
.7
2792
.7
2292
.4
2008
.5
2842
.9
2242
.4
2893
.1
1391
.3
2543
.4
2643
.4
1341
.3
1441
.4
2743
.5
2443
.3
1291
.3
2843
.6
1491
.4
1241
.2
2343
.3
2943
.6
1541
.4
1191
.2
B1
B2
B3
B4
n=55n=42
n=55
n=44
[M+K]+
[M+2K]2+
[M+2K]2+[M+3K]3+
5SUMITOMO KAGAKU 2008-II
Investigation of Synthetic Polymer Characterization by means of SEC/ESIMS
cannot be determined. Since the resolution of a stan-
dard q-TOF mass spectrometer is approximately 10,000
(FWHM) even with precise adjustment, when using
isotope peak intervals, the charge numbers can be
determined only up to 4 at m/z=2,000, and up to 2 at
m/z=4,000. Thus, it is necessary to use FT-MS with
higher resolution or find other methods.
Estimating Charge Number Using Molecular
Weights of Constitutional Repeating Units
Fig. 9 shows an enlarged view of spectrum B2 (Fig.
6). In spectrum B2, the peaks are aligned from the low
m/z side (the left side) in the order of the higher
charge number 5,4,3 and 2. At this stage, because the
charge numbers of each peak is still unknown, they are
represented as +(z+1)K, +zK, +(z-1)K. a) and b) show
enlarged spectra of the parts indicated by the dotted
lines. Because there is not adequate mass resolution,
the isotopic peak intervals cannot be read. The figures
indicate the intervals between the peaks x1 to x8 and
those between the peaks y1 to y6. These intervals are
derived from the constitutional repeating unit (CRU) of
PMMA. Therefore, they match the molecular weight in
the CRU, and when measuring a 1+ charged ion, the
value is 100Da. However, as multiply charged ions are
being measured in the diagram, the values are smaller
than 100Da. In the previous section, it was explained
that the detected m/z value decreases as the charge
number increases, but in the same way, smaller molec-
ular weights are also detected in the CRU -- such as
be determined. In b) of Fig. 7, the resolution is approxi-
mately 2,000 (FWHM). Therefore, in order to measure
the molecular weight of synthetic polymers using the
SEC/ESIMS method, it is necessary to use a high reso-
lution mass spectrometer, such as TOFMS or TF-
MS.12)
Fig. 8 shows the mass spectra of the same sample
containing the same components, but measured using
a quadrupole time-of –flight (q-TOF) mass spectrome-
ter (q-tof Ultima of Micromass Ltd.). Although the
details of the measurement conditions are omitted
here, the mass resolution is approximately 6,000
(FWHM) for actual measurements. As the diagram
clearly shows, it is easy to determine 2+ charged ions.
However, even at a resolution of 6,000, 3+ charged ions
Fig. 9 Expanded Spectrum A2 14.3 × 7 = 100.1 16.7 × 6 = 100.2 z = 7
m/z1000 1250 1500 1750 2000 2250 2500 2750 30000
20406080
100B2
2041
.7
2055
.9
2084
.5
2099
.0
2027
.6
2113
.1
2127
.4
2070
.4
14.1 14.2 14.5 14.1 14.5 14.1 14.3
+zK +(z-1)K
+(z+1)K +(z-2)K
x1
2442
.2
2425
.4
2459
.0
2475
.3
2492
.2
2509
.1
16.8 16.8 16.3 16.9 16.9
x2 x3x4
x5 x6 x7 x8
y1 y2 y3y4 y5 y6
a)
b)
Fig. 7 Profile Spectrum of n=27 peek (see Fig.6)
m/z2930 2935 2940 2945 2950
m/z1480 1485 1490 1495 1500
δ δ δ
δδ
δ = 1Da
0
10
2030
40
50
0
20
40
60
80
100
2943
.6
2944
.6
2945
.6
2942
.6
2946
.6
1491
.4
a) n=27,[M+K]+
b) n=27,[M+2K]2+
Fig. 8 High resolution SEC/ESIMS Spectrumresolution = 6000(FWHM)
δ1
δ2
δ1
δ1
δ1
a) n=27,[M+K]+
b) n=27,[M+2K]2+
c) n=55,[M+3K]3+
m/z
100
%
2943.6
2942.6
2944.6
2945.6
m/z
0
100
%
0
100
%
0
1491.3
1490.8
1491.8
1492.3
1492.8
m/z1928 1930 1935 1940 1945 1950 1953
1478 1480 1485 1490 1495 1500 1503
2930 2935 2940 2945 2950 2955
1941.8
δ1 = 1Da
δ2 = 0.5Da
6SUMITOMO KAGAKU 2008-II
Investigation of Synthetic Polymer Characterization by means of SEC/ESIMS
1/2, 1/3 and so on -- as the detected charge number
increases. The reason why the intervals on the high
molecular weight side shown in a) are greater than
those on the low molecular weight side shown in b) is
because each peak on the high molecular weight side
in b) is one charge number smaller than that on the
low molecular weight side in a). The average of the
+zK peak intervals is 14.3 and that of the +(z-1)K peak
intervals is 16.7. As charge numbers of these peaks are
equivalent to z and z-1, the following equation can be
established:
z × 14.3 ≈ (z -1) × 16.7
As this contains inaccuracies, it cannot be strictly
“=.” However, if making z=7, the following equations
can be obtained:
7 × 14.3 = 100.1 6 × 16.7 = 100.2
Therefore, the charge number and molecular weight
in the CRU can be estimated. In this case, the molecu-
lar weight in the CRU can be estimated as 100, and this
is certainly consistent with PMMA.
Naturally, because this method can be applied only
to cases in which the selected peaks are those of the
same kind of synthetic polymer, it is necessary to take
extra caution when applying the method to mixed poly-
mers. From our experience however, it can be said that
this molecular weight estimation method is highly
effective in many situations.
Mass Spectra and Molecular Weights of Com-
ponents Having High Molecular Weights
Once the charge number has been determined, the
molecular weight can be readily estimated. However,
when comparing with theoretical values, mass spectra
of components with high molecular weights have fur-
ther problems.
Fig. 10 shows the simulation results of M+ mass
spectra of 10-mer a) and 100-mer b) of PMMA. The
mass resolution is 10,000 (FWHM). However, it should
be noted that it is not strictly M+, as electron mass is
not taken into account. (10-mer and 100-mer refer to
the value “n” described in Fig. 2.)
As previously mentioned, the mass spectrum reflects
the abundance ratios of stable isotopes. Because the
molecular weight of the 10-mer is approximately 1,000
and the mass resolution is 10,000, the half-value width
of the isotopic peak is around 0.1Da. Therefore, ade-
quate separation and extremely sharp peaks can be
achieved. However, peaks on the low molecular weight
side are produced by molecules composed of a single
isotope, and are called a mono-isotopic peaks. The
value of such peaks is equivalent to the exact mass, and
is used when conducting accurate mass measure-
ments. Roughly speaking, this exact mass can be used
for estimating low molecular weight components with
molecular weights of up to approximately 2,000.
However, as the molecular weight approaches 10,000
in the 100-mer, the appearance of the spectrum sud-
denly changes. Even though the mass resolution is
10,000, the molecular weight is also 10,000. Therefore,
the half-value width of each isotopic peak becomes
1Da, thus allowing the peak width to expand to 10
times wider than that of the 10-mer. Also, because the
number of atoms constituting the molecule is large,
there are more varieties of isotope combinations, thus
increasing the number of peaks. For this reason, as
shown in b) of Fig. 10, each isotopic peak can be barely
separated, thus shaping the peak like a broad moun-
tain. In addition, as the mono-isotopic peak appears at
the foot of this broad peak, its peak cannot be identi-
fied. In other words, unless precise observation is con-
ducted to define which area of the peak is considered
as the molecular weight of the 100-mer, the molecular
weights obtained from the experiment cannot be accu-
rately compared. Although it is reasonable to regard
10213.3828 at the peak top as the molecular weight
measured using MS, because this value differs from
Fig. 10 Simulated PMMA oligomer Peak(M+) resolution = 10000(FWHM)
a) 10-mer b) 100-mer
C60H98O24
Exact Mass : 1202.6448Mol. Wt. : 1203.4057
C510H818O204
Exact Mass : 10207.3635Mol. Wt. : 10213.8295
Mono isotopic peak10207.3635
mass1194 1196 1198 1200 1202 1204 1206 1208 1210 1212 1214 1216
mass10230102251022010215102101020510200
Mono isotopic peak
a)
b)
0
100
%
0
100
%
1202.6448
1203.6448
1204.6526
10213.3828
OO
OO
OO
10
OO
OO
OO
100
Synthetic Polymer Mixtures
Figs. 11–13 show the measurement results of a
PPG/PMMA mixture. Needless to say, these two com-
ponents cannot be distinguished using a UV chro-
matogram or TIC, neither can the presence of the two
components be presumed. Although precisely analyz-
ing each peak using the spectra shown in Fig. 12 may
indicate that it is a PPG/PMMA mixture, it requires
time, effort and creativity. When only using an MS
spectrum for analysis, as soon as samples become mix-
tures, analysis becomes extremely difficult.
Therefore, we attempted the analysis via a two-
dimension plot (Fig. 13). In the diagram, the data
obtained from the SEC/ESIMS measurement are
the molecular weight (MW), when measuring and eval-
uating the molecular weight of a component having
high molecular weight using a mass spectrometer, it is
necessary to confirm the detected molecular weight
through the spectrum simulation method introduced
here. In this paper, the value at the peak top of the sim-
ulation spectrum was used as the theoretical molecular
weight and then compared to the actual measured
value using MS.
Accuracy of Molecular Weight Estimation
Using Multiply-charged Ion Peaks
Table 1 shows the results of estimates of the molec-
ular weights of the x5 and y5 components shown in
Fig. 9. In the calculated values, MW represents molec-
ular weights, which are shown for reference purposes.
Using a q-MS, components having molecular weights
of 14,000 or more were estimated from +6K and +7K
ions. As a result, it has been confirmed that the error is
within ±1Da. Therefore, the error is within 100ppm,
meaning an accuracy of ±0.1 in the m/z=1,000 conver-
sion. This is extremely good consistency, and we there-
fore believe that this method can be used for structural
analysis of actual synthetic polymers. According to our
investigation, molecular weights of up to approximately
18,000 can be accurately estimated with +8K ions when
using PMMA.
For reference purposes, the m/z detection accuracy
using a q-MS is usually ±0.1Da. For example, if the
detection accuracy for 6+ charged ions is ±0.1Da, the
error in the molecular weight obtained from the calcu-
lation becomes ±0.6Da. Therefore, theoretically, the
accuracy of molecular weight estimation from multiply
charged ions using a q-MS is acceptable. If the m/z
detection accuracy is improved by one order of magni-
tude, to approximately ±0.01Da, we believe that the
accuracy of molecular weight determination for syn-
thetic polymers will greatly improve, thus enabling
more detailed structural analysis by using a combina-
tion of a q-MS and SEC/ESIMS.
7SUMITOMO KAGAKU 2008-II
Investigation of Synthetic Polymer Characterization by means of SEC/ESIMS
Table 1 PMMA MW from SEC/ESIMS Spectrum
2084.52492.2
Peak(m/z)
M+7KM+6K
ion type
14718.614317.8
ExperimentalMW
14718.714318.4
Peak topCalculated
14719.014318.6
MW
Fig. 11 Size-exclusion Chromatograms of PMMA+PPG
a) UV=220nm b) TIC(+KCl)C1~C3 ; averaging area of the mass spectrum in Fig.12Column ; Tsk-gel Super HZ (2500+2000*2+1000)
a)
b)C1
min27.5 30 32.5 35 37.5 40 42.5
mAU
1234567
min27.5 30 32.5 35 37.5 40 42.50
200000400000600000800000
1000000120000014000001600000
C2 C3
Fig. 12 SEC/ESIMS(+KCl) Spectrum at C1~C3
Sample ; PMMA+PPG
C1
C2
C3
m/z500 750 1000 1250 1500 1750 2000 22500
20
40
60
80
100
m/z500 750 1000 1250 1500 1750 2000 22500
20
40
60
80
100
m/z750 1000 1250 1500 1750 2000 22500
20
40
60
80
100
expressed as follows: the vertical axis represents m/z ;
the horizontal axis represents dissolution time (min-
utes); and the intensity gradation represents the
strength of each ion peak. The presence of the two
components can be presumed at a glance. Analyzing
the mass spectrum based on this information, it is clear
which peak groups correspond to PPG, and which to
PMMA, from among the components. In addition, with
respect to charge state, even the charge state of high
molecular weight components having complex mass
spectra (and therefore being difficult to have their
charge state determined) can be estimated from the
elution behavior of each component. From Fig. 13, it
can be observed that the separation of high molecular
weight components is not adequate when using SEC.
Furthermore, it can be readily understood that in order
to analyze high molecular weight components more
precisely, improving the separation capability can be
effective by increasing the number of columns, and by
other methods.
Thus, it is obvious that 2-D plots are extremely effec-
tive in order to analyze synthetic polymers having com-
plex compositions, as well as mixtures.
GB/MA Copolymer
Figs. 14–16 show the measurement results of a
GB/MA copolymer, which is a more practical sample.
8SUMITOMO KAGAKU 2008-II
Investigation of Synthetic Polymer Characterization by means of SEC/ESIMS
Fig. 15 GB/MA copolymer SEC/ESIMS Spectrums at D
COC
R2
O O
R1
O
O
O
G M
149
GB MA
–149*3
1256.5(4, 2)
876.3792.2 960.3
b) Fr=400V –149*2
a) Fr=200V–149*4
(3, 3)1320.7
1384.7 1618.9
1683.8
(2, 4) (2, 5)
(1, 6)
020406080
100
m/z0 250 500 750 1000 1250 1500 1750
m/z0 250 500 750 1000 1250 1500 1750
020406080
100
a) Fragment Voltage=200V b) Fragment Voltage=400V
Fig. 13 2-D Plots of SEC/ESIMS of PMMA+PPG
Column ; Tsk-gel Super HZ (2500+2000*2+1000)
28 30 32 34 36 38 40 42 44
400
600
800
1000
1200
1400
1600
1800
2000
2200
2400
2600
2800 m/zm/z
minmin
PMMA+PMMA+PPGPPG
PMMAPMMA
PPGPPG
+K+K+2K+2K
+3K+3K
+4K+4K
+6K+6K
+7K+7K
+K+K+2K+2K+3K+3K
Fig. 14 SEC of GB/MA copolymer
mAU
–500
50
min20 25 30 35 40 450
200000
400000
600000
a) 210nm
b) TIC
D
a) 210nm b) TIC of SEC/ESIMSD ; averaging area of mass spectrums in Fig.15Column ; Tsk-gel Super HZ (2500+2000*2+1000)
9SUMITOMO KAGAKU 2008-II
Investigation of Synthetic Polymer Characterization by means of SEC/ESIMS
This sample was synthesized by radical polymerization.
The weight-average molecular weight (Mw) is ≈ 10,000
and the distribution (Mw/Mn) is ≈ 1.9 (indicated as
polystyrene). Although the Mw is rather small, the
sample has a wide Mw/Mn and a random sequence,
thus making its composition extremely complex. For
this reason, with ESIMS measurement using the flow
injection method that does not utilize SEC, it was diffi-
cult to determine whether or not the sample had been
ionized because the spectrum was extremely compli-
cated.
With the SEC/ESIMS (+KCl) method, the mass
spectrum that reflects the copolymer composition was
detected as a [M+K]+, thus making it easy to estimate
the molecular weight for low molecular weight compo-
nents (up to approximately MW=2,500). As a result,
the composition of each peak could be identified (Fig.
15, a)). It was also understood that more precise infor-
mation can be obtained through the capillary skimmer
CID method (Fig. 15, b)).
In addition, it was understood that information
regarding dissolution behavior and charge state, which
is useful for analyzing mass spectra, can be obtained
from 2-D plots (Fig. 16). In the 2-D plot, it was also
clearly confirmed that a 2+ charged ion that could not
have been identified in a regular mass spectrum was
also detected. However, because the spectrum of this
2+ charged ion was extremely complex, it was difficult
to specify the components and estimate the molecular
weight.
Fig. 17 shows the complexity of the mass spectrum
of a copolymer. We believe that from this model spec-
trum, it can be observed that although it is a mere A/B
copolymer, its mass spectrum can become unimagin-
ably complex. In order to analyze a complex copolymer
which has several (>3) kinds of constitutional repeating
unit, it is essential to use chromatography which has
high separation capability.
Fig. 16 GB/MA copolymer 2–D plot by SEC/ESIMS (1,6) means G=1, M=6
35 40 45 50 55 60 65
400
600
800
1000
1200
1400
1600
1800
2000
2200
2400
m/zm/z
minmin
(1, 6)
(3, 4)
(1, 5)(2, 4)
(1, 4)
+K+K+2K+2K
COC
R2
O OR1
O
O
OG M
(2, 3)
Fig. 17 Complexity of copolymer mass spectrum
α=2
α=3
α=4α=5
α
α β
(A)
(0,2)
(2,0)
(1,1)
(0,5)
(5,0) (2,3)
(A) (B)
a)
b)
m/z
m/z
a) homo-polymer model b) co-polymer model α, β=1,2,3 · · ·, (2,3) means α=2, β=3
10SUMITOMO KAGAKU 2008-II
Investigation of Synthetic Polymer Characterization by means of SEC/ESIMS
Conclusion
From the results described above, the following fact
has been confirmed: although hydrophobic synthetic
polymers can be ionized up to molecular weights of
tens of thousands by the SEC/ESIMS method using a
cationization agent, if the composition of the polymer is
complex, it is hard to estimate its molecular weight.
Therefore, it can be concluded that in order to conduct
structural analysis of a synthetic polymer, it is more
important to use chromatography which has high sepa-
ration capability, rather than a high performance mass
spectrometer.
Furthermore, using the PMMA in Fig. 2 as an exam-
ple, when analyzing the molecular weight of the CRU,
if only n=100 compounds are extracted, refined and
measured, only the molecular weight of the n=100 com-
pounds can be clarified, not that of the CRU. In order
to obtain the molecular weight of the CRU, a group
having consecutive numbers such as n = -- - - 98, 99,
100 --- must be analyzed. Such groups cannot be ana-
lyzed using UV or TIC, but the 2-D plots shown in Figs.
13 and 15 are effective. We believe that if there is a
function which can plot only the trivalent ion spectrum
directly from a 2-D plot, and which can conduct back-
ground processing, mass spectra which are easier to
analyze can be achieved, resulting in more efficient and
precise analysis.
In recent years, there have been some reports on
new software for copolymer composition analysis.13) Its
development is expected to continue in the future.
Acknowledgements
Lastly, we would like to express our deep apprecia-
tion to Professor Ryuichi Arakawa, Faculty of Chem-
istry, Materials and Bioengineering, Kansai University,
for showing his interest in this research and providing
us with valuable advice and encouragement.
References
1) S. Asano, K. Yamamoto, O, Hiroaki, M, Kato, SUM-
ITOMO KAGAKU, 1998-!, 68 (1998).
2) X. Michael Liu, E. Peter Maziarz, David J. Heiler
and George L. Grobe, J.Am.Soc.Mass Spectrom., 14,
195 (2003).
3) T. Kashiwagi, O. Hiroaki, M. Kato, H. Takigawa,
SUMITOMO KAGAKU, 1993-!!, 71 (1993).
4) John B. Fenn, Matthias Mann, Chin Kai Meng,
Shek Fu Wong and Craige M. Whitehouse, Mass
Spec. Revi., 9, 37 (1990).
5) Ryuichi Arakawa, Syouji Okuno, Bunseki, 2002, 9,
502 (2002).
6) Laszlo Prokai and William J. Simonsick, Jr., Rapid
Commun. Mass Spectrom., 7, 853 (1993).
7) Michel W. F. Nielen, Rapid Commun. Mass Spec-
trom., 10, 1652 (1996).
8) Y. Tsuchida, K. Yamamoto, H. Yamada, S. Morimo-
to, K. Yamamoto, 54th Annual conference on Mass
Spectrometry Osaka Japan, 42 (2006).
9) Y. Tsuchida, K. Yamamoto, 9th National Symposium
on Polymer Analysis and Characterization, 147
(2006).
10) Y. Tsuchida, K. Yamamoto, H. Yamada, S. Morimo-
to, K. Yamamoto, 11th National Symposium on Poly-
mer Analysis and Characterization, 71 (2006).
11) Y. Yokoyama, J.Am.Soc.Mass Spectrom., 18, 1914
(2007).
12) David J. Aaserud, Laszlo Prokai and William J.
Simonsick, Jr., Anal.Chem., 71, 4793 (1999).
13) Steffen M. Weidner, Jana Falkenhagen, Sergey
Maltsev, Volker Sauerland and Marian Rinken,
Rapid Commun. Mass Spectrom., 21, 2750 (2007).
11SUMITOMO KAGAKU 2008-II
Investigation of Synthetic Polymer Characterization by means of SEC/ESIMS
P R O F I L E
Yoshinobu TSUCHIDA
Sumitomo Chemical Co., Ltd.IT-Related Chemicals Research LaboratorySenior Research Associate
Hiromi YAMADA
Sumitomo Chemical Co., Ltd.IT-Related Chemicals Research LaboratoryResearcher
Keiko YAMAMOTO
Sumitomo Chemical Co., Ltd.IT-Related Chemicals Research LaboratoryResearch Associate