10
Then for an unknown the data set {[
η
]
i
,
V
i
} is also measured.
In trin s ic V isc o s ity R e s u lts
9 .7
1 1 .9
1 4 .1
1 6 .2
1 8 .4
2 0 .5
-1 .3 3
-0 .9 1
-0 .4 8
-0 .0 5
0 .3 7
0 .8 0
R eten tion V olu m e (m l)
Figure 2 Typical Plot of Intrinsic Viscosity versus Retention Volume for Broad Sample
It is clear that the data set {
M
i
,
V
i
} for the unknown sample can be mathematically extracted by
dividing the data of
Figure 2
into that of
Figure 1
. The full data set {
C
i
, [
η
]
i
,
M
i
,
V
i
} for the
unknown sample could then be constructed, the first and third terms of which constitute the
desired MWD.
Universal Calibration is the best means of column calibration in SEC. It is applicable to
a wide range of polymers, the polymer standards do not have to be the same type as the
unknown, and it applies to copolymers as well as to both linear and branched samples.
However, it necessarily assumes that the mechanism of retention on the columns for both
sample and standards is pure size exclusion.
(3) SEC
3
The molecular weight distribution can be obtained directly from an on-line light scattering
detector connected in series before the refractometer and viscometer detectors. This is the
SEC
3
system. The data set {
C
i
, [
η
]
i
,
M
i
,
V
i
} is obtained just as with Universal Calibration.
However, the
M
i
values are calculated directly from the LS detector signal, not derived through
the column calibration. In addition to the convenience of not having to run calibration standards,
SEC
3
offers the real advantage of not being constrained by the size exclusion separation
mechanism.
MOLECULAR SIZE AND BRANCHING
Regardless of how one determines MWD, whether by Universal Calibration or Triple
Detection, the viscometer detector is useful for determining the size of molecules in solution.
The molecular size can then be related to the secondary structure of the polymer, either in
terms of chain stiffness, conformation, or branching.
Molecular size is commonly computed as the radius of gyration,
R
g
, which can be
determined rigorously only from the initial slope of multi-angle light scattering measurements.
Unfortunately, the multi-angle slope approach has only a limited range of applicability, as the
Log
IV
