Fractionation of Industrial Starch Polysaccharides by Field-Flow Fractionation
Justin R. Engle and S. Kim R. Williams
Department of Chemistry and Geochemistry, Colorado School of Mines, Golden, CO 80401, krwillia@mines.edu
Objectives
Experimental
¾Determine best sample preparation techniques for the dissolution
of starch in organic media
Starch samples with varying amylose content (0, 25, 50, 70%) were
provided by National Starch and Chemical Company (Bridgewater, NJ).
¾Perform fractionation of amylose and amylopectin components of
industrially important starch samples using thermal field-flow
fractionation (ThFFF)
Carrier liquid was distilled dimethyl sulfoxide (DMSO), flow rate of 0.1
mL/min. Temperature programming was used: initial ΔT=80ºC for 10
minutes then decayed to ΔT=10ºC.
¾Determine molecular weight (MW) for amylopectin
Sample Preparation
Background
Method 1: starch cooked in distilled dimethyl sulfoxide (DMSO) in an oil
bath at ~80ºC for 3 days with constant stirring. Sample concentrations
were 1.0 mg/mL.
OH
n
OH
Amylose
O
OH
OH
OH
CH2OH
OH
OH
6
O
HO CH OH
2
O
OH
O
OH
OH
Amylopectin
Characterization of starch has previously been performed using
size-exclusion chromatography (SEC). The use of SEC for
polysaccharides is limited due to low exclusion limits of the packed
column and shear degradation of the ultrahigh molecular weight
components. An alternative to SEC is field-flow fractionation (FFF). In
1994, Lou et al. reported the use of thermal FFF (ThFFF) to perform
the separation of pullulan with varying molecular weight and of corn
starch with varying amylose and amylopectin content [2]. ThFFF uses
a temperature field applied perpendicular to the carrier liquid flow to
promote the retention and fractionation of analytes
based on the ratio of thermal diffusion coefficient DT to normal diffusion
coefficient D. The retention time tr of the analyte is therefore controlled
by the temperature difference across the channel ΔT, the size and
molecular weight of the analyte, and parameters that affect DT.
D
t0
λ=
≈
DT ΔT 6tr
3.0x10
7
4
2.5x10
7
2.0x10
7
1.5x10
2
7
1.0x10
6
t0
0
0
5.0x10
10
20
30
40
50
Molar mass signal
1.4
1.2
2.0x10
6
1.5x10
6
1.0x10
6
5.0x10
5
1.0
0.8
0.6
0.4
t0=4.5 min
0.2
0.0
10
20
30
40
50
6
0.0
10
Observations for Cooking Method 1
¾Figure 1 light scattering trace shows sample retention
MW increases with increasing elution time
Measured MW range from 5.0x106 to 3.5x107 Da
¾Figure 2 has similar light scattering trace as Figure 1
Sample retained from 20 minutes to 60 minutes – likely that of
amylopectin
MW initially increases with elution time
MW range of 5.0x105 to 1.7x106 Da
Range is significantly lower than literature (107 Da)
Model T-101 thermal channel (cover removed).
20
30
40
50
60
90 degree LS Signal (volts)
Shorter retention of sample (from ~4 minutes to 45 minutes) than seen
in cooking Method 1
Range 1x106 to 8x106 Da assuming amyloopectin elution at 20 min
Values lower than literature (107 Da)
Molar mass signal follows I90 trace
Sample (prep
method)
Figure 2 – Fractogram of 50% amylose and 50% amylopectin with molecular
weight overlay. Sample prepared by Method 1.
Literature values for amylopectin from plant sources is ~107 Da
On-line Detection
Dawn EOS multi-angle light scattering (MALS), Optilab DSP
differential
refractometer (dRI), and quasi-elastic light scattering (QELS) from
Wyatt Technologies, Santa Barbara, CA.
Data Analysis
Berry fitting method used for molecular weight and root mean squared
(rms) radius calculations
dn/dc values for amylose 0.0659 and amylopectin 0.066 in DMSO [6]
2.0x10
2
MW
(g/mol)
0% amylose (1)
100% amylopectin
--(1.6±0.2)x107
50% amylose (1)
50% amylopectin
--(1.2±0.1)x106
50% amylose (2)
50% amylopectin
--(3.6±0.4)x106
0.0
60
Sample retained between 20 to 50 minutes is likely amylopectin
Thermal channel: Model T-101 (Postnova Analytics, Salt Lake City, UT)
Length: 32 cm Breadth: 2 cm Thickness: 0.0127 cm
6
t0=4.5 min
Figure 1 – Fractogram of 100% amylopectin with molecular weight overlay.
Sample prepared by Method 1.
Time (min)
Instrumentation
4.0x10
4
Table1 – Calculated molecular weight averages for amylopectin in Figures 1-3
from elution time of 20 minutes and onwards.
0.0
60
Time (min)
0
In the past decade, flow FFF has been in the spotlight with respect to
polysaccharides analyses [3-5] using an aqueous carrier liquid.
Separation of polysaccharides of starches by flow FFF is based solely
on the hydrodynamic size of the macromolecule whereas ThFFF has
an additional composition sensitivity.
6
¾Figure 3 fractogram significantly different than that of fractograms
obtained by cooking Method 1 (Figures 1 and 2)
7
OH
OH
6.0x10
7
3.5x10
OH
6
Figure 3 – Fractogram of 50% amylose and 50% amylopectin with molecular weight
overlay. Sample prepared by Method 2
O
n
6
Observations for Cooking Method 2
Molar mass signal
CH2OH
O
O
OH
8.0x10
Time (min)
Molecular weight (g/mol)
O
Side chain O
8
Molecular Weight (g/mol)
O
OH
OH
O
I90 signal (volts)
OH
CH2OH
O
7
0
Results and Discussion for Cooking Method 1
I90 Signal (volts)
CH2OH
1.0x10
0
Method 2: starch cooked in 90% aq. DMSO in a boiling water bath (9496ºC) for 1 hour with constant stirring. Sample was extracted when
temperature was near room temperature. Sample concentration 2.0
mg/mL.
2
CH2OH
Molar mass signal
10
Conclusions
¾Both cooking methods allowed complete dissolution of starch
¾Cooking Methods 1 and 2 yielded different elution profiles
¾Both cooking methods generally led to lower MWs for amylopectin
(assuming elution after 20 minutes) than literature value (107 Da)
¾Weak dRI signals may be contributing to underestimation of Mws
Future work
¾Collection of amylose and amylopectin via organic flow field-flow
fractionation to use as size standards
¾ThFFF-MALS-dRI for quantification of amylose and amylopectin
content
Comparison of amylose and amylopectin content using standard
iodine potentiometric titration method to determine accuracy of
quantification of amylose and amylopectin content.
¾Calculation of thermal diffusion and molecular (normal) diffusion
based on retention times
References
Acknowledgments
NSF-CHE-0515521
Dr. Tom Hahn (National Starch)
Dr. Claudia Lohmann (Colorado School of Mines)
Dr. Sohyel Tadjiki (Postnova Analytics)
J. Ray Runyon (Colorado School of Mines)
Molecular Weight (g/mol)
Starches are important macromolecules with many industrial
applications, particularly in the food and pharmaceutical industries.
Their uses range from gelling agents to drug transport and even
energy storage. Properties of starches vary based on the plant or
bacterial source and are connected to the amylose and amylopectin
content. Amylose is a linear polysaccharide with α-1,4 glycoside
bonds and has a molecular weight around 106 Daltons. Amylopectin is
a highly branched polysaccharide with α-1,6 glycoside bonds along a
α-1,4 glycoside bond chain backbone and has a molecular weight
CH OH
around 107 Daltons [1].
O
Results and Discussion for Cooking Method 2
[1] Richardson, S.; Gorton, L. Anal. Chim. Acta. 497, 27-65 (2003).
[2] Lou, J.; Myers, M.N.; Giddings, J.C. J. Liq. Chromatogr. 17, 3239-3260
(1994).
[3] Wittgren, B.; Wahlund, K.-G.; Andersson, M.; Arfvidsson, C. Int. J. Polym.
Anal. Charact. 7, 19-40 (2002).
[4] Roger, P.; Baud, B.; Colonna, P. J. Chromatogr A. 917, 179-185 (2001).
[5] Stevenson, S.G.; You, S.; Izydorczyk, M.S.; Preston, K.R. J. Liquid
Chromatogr. 26, 2771-2781 (2003).
[6] Polymer Handbook 4th Ed.. J. Brandrup, E.H. Immergut, E.A. Grulke Eds.
John Wiley & Sons, Inc. New York: 1999.