Macro-ion Characterization - ACS Publications - American Chemical

Chapter 10

Molecular-Weight Distributions of WaterSoluble Polyelectrolytes Gel Permeation Chromatography Spectrophotometry—Low-Angle Laser Light Scattering

Downloaded by UNIV LAVAL on June 7, 2017 | http://pubs.acs.org Publication Date: December 13, 1993 | doi: 10.1021/bk-1994-0548.ch010

Ellen M. Meyer and Stephen R. Vasconcellos Betz Industrial, 4636 Somerton Road, Trevose, PA 19053-6783

Recent developments in the use of gel permeation chromatography (GPC) for the determination of the molecular weight distributions of water soluble cationic polyelectrolytes are discussed. Absolute molecular weight measurements were made using a low-angle laser light scattering (LALLS) spectrophotometer. The polymers studied included five samples of an epichlorohydrin dimethylamine (epi/dma) copolymer with weight average molecular weights ranging from 103 to 10 , and four samples of polymeric diallyldimethylammonium chloride (dadmac) with weight average molecular weights ranging from 10 to 10 . These polymers are typically used for water clarification in waste treatment applications. 5

4

5

A correlation exists between the molecular weight and activity of polymers in wastewater treatment applications, but there is insufficient data to determine the effect of the molecular weight distribution on performance. Although the techniques used to determine the molecular weight distribution of polymers have been established for some time, the majority of the research has been conducted on non-aqueous polymers. Since most of the polymers used for wastewater treatment are water soluble, and often contain anionic or cationic functionalities, this research was an attempt to improve the database of molecular weight distribution information available for this important class of polymers. The cationic polymers, epichlorohydrin/dimethylamine (epi/dma) and diallyldimethylammonium chloride (dadmac), are well known for their ability to clarify wastewater (I). Previous work with these polymers has been performed with gel permeation chromatography (GPC) using calibration standards chemically different from the polymers being characterized (Bauer, D.S., Betz Chromatography Services, unpublished data). In the present study, GPC was performed in

0097-6156/94/0548-0131$06.00/0 © 1994 American Chemical Society

Schmitz; Macro-ion Characterization ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

132

MACRO-ION CHARACTERIZATION

conjunction with low angle laser light scattering ( L A L L S ) to obtain absolute molecular weight distributions.


Experimental Differential refractive indices (dn/dc) were measured with the L D C / M i l t o n R o y Chromatix K M X - 1 6 differential refractometer in the same solvent system as that used for the L A L L S experiments. Four different concentrations o f polymer were used and the differential refractive index values were averaged. If the dn/dc values changed with concentration, the y-intercept o f a plot o f dn/dc vs concentration was used in the molecular weight calculations. Preliminary to running the G P C / L A L L S experiments, the weight average molecular weights o f the polymers were measured by using static L A L L S in 1 M N a C l . These experiments were conducted using the Chromatix K M X - 6 instrument from L D C / M i l t o n Roy. The system parameters are contained in Table I. T w o o f the polymer samples were also measured in the 0.2 M L i N O 3 / 0 . 1 % T F A solvent system used for the G P C / L A L L S experiment. The G P C / L A L L S instrumentation configuration consisted o f an L D C Analytical ConstaMetric 3200 B i o Solvent Delivery System, Rheodyne M o d e l 7125 Syringe Loading Sample Injector, L D C / M i l t o n Roy Chromatix K M X - 6 L o w Angle Light Scattering Photometer, and L D C Analytical RefractoMonitor I V Refractive Index Detector. Data were collected and processed with the use o f the L D C Analytical P C L A L L S software package. Synchropak C A T S E C columns were used in the following series: C A T S E C 300 A (50 mm χ 7.8 mm) guard column, C A T S E C 4000 A (300 mm χ 7.8 mm) column, C A T S E C 300 A (30 mm χ 7.8 mm) column. The injector sample loop volume was calibrated by injecting a known concentration o f sodium benzoate into a volumetric flask, filling to the mark, and measuring the U V - V I S absorbance. The absorbance was compared to a calibration curve to obtain the volume o f injected sodium benzoate. The interdetector volume was calibrated by measuring the time difference between the L A L L S and R I detector responses for an unfractionated polymer. The pump was calibrated daily by measuring the time required to fill a volumetric flask. In order to avoid complications from impurities in the sample, the dadmac samples were dialyzed to remove any residual synthesis process salts. The solvent systems were prepared using doubly distilled water. For the G P C / L A L L S experiments, a mobile phase o f 0.1% trifluoroacetic acid ( T F A ) and 0.2 M was chosen (Bauer, D . S . , Betz Chromatography Services, unpublished data.). D . J. Nagy et al. have shown universal calibration behavior with cationic polymers in a similar solvent system (0.1% TFA/0.20 Ν N a N 0 3 ) using C A T S E C columns (2). The intrinsic viscosities were measured by the Chromatography Services Group at Betz using the Schott-Gerate Automatic Viscometer.

L1NO3

Results Epi/dma Polymers. Table II summarizes the numerical results for the epi/dma polymers. From Table II, it is evident that the M w values obtained from the G P C / L A L L S experiments with 0.2 M L i N O 3 / 0 . 1 % T F A , are consistently lower than


Schmitz; Macro-ion Characterization ACS Symposium Series; American Chemical Society: Washington, DC, 1993. Mz l,810,000(+/-280,000) 131,000(+/-17,000) 41,100(+/-8,200) 15,000(+/-2,200) 3,280(+/-650)

GPC L A L L S Mw 235,000(+/-20,000) 36,100(+/-3,200) 19,700(+/-2,200) 4,180(+/-420) 2,350(+/-590)

Static L A L L S Mw

324,000(+/-4,000) 43,800(+/-500) 30,100(+/-400) 8,800(+/-600) 3,510(+/-20)

0.22 0.16 0.13 0.05 0.03

A Β C D Ε

Table Π. Epi/dma Results

1.7703e-8 0.15 6-7 degrees 4.93 mm 0.2MLiNO3/0.1%TFA 1.335 4.865 877.36 57 microliters - 0 . 6 mL/min 60.5 microliters

Intrinsic Viscosity

~0.2 mL/min

1.7703e-8 0.15 6-7 degrees 15 mm l.OMNaCl 1.343 4.835 881.33

GPC/LALLS

Polymer Sample

Attenuator Constant Fieldstop Annulus Cell Length Solvent Refractive Index Scattering Angle l/(sigmaT) Interdetector Volume Flow L o o p Volume

Static L A L L S

Table I. LALLS System Parameters


15,300(+/-980) 10,800(+/-950) 6,900(+/-2,600) l,880(+/-220) l,450(+/-520)

Mn

1

S.

f to

3

f

ζ/2

r f ο

Μ

ο

ο

η

©

134


the values obtained using static L A L L S with 1 M N a C l . Since the hydrodynamic volume o f a polymer may be influenced by the ionic strength o f the solvent (3,4), it is not surprising that the values obtained from these two experiments should differ. Subsequent static L A L L S measurements using 0.2 M T F A indicate that lower weight average molecular weight values are obtained with this solvent system compared to the N a C l solvent system (Table III). With the same solvent system, the G P C / L A L L S and static L A L L S values are within experimental error. Figures 1-5 show the unusual curve shapes seen for some o f these polymers. The L A L L S curve o f Polymer A seems to indicate that the polymer is being excluded from the column pores, but the sharp rise in the L A L L S peak is not duplicated in the R I peak. Polymers D and Ε appear to be mixtures o f different low molecular weight batches. Polymer D has three distinct peaks in the L A L L S curve and a shoulder on the R I curve. The L A L L S data o f Polymer Ε show one large peak and another small peak with a much higher molecular weight. The R I data does not show this extra peak, possibly because the concentration is too low. Both Polymers D and Ε are difficult to measure using this method because the L A L L S technique is not very sensitive to low molecular weight polymers. Polymers Β and C have nearly gaussian curve shapes.


L1NO3/0.1%

D a d m a c Polymers. Table I V summarizes the results for the dadmac polymers. Figures 6-9 show typical chromatograms for these samples. A s with the epi/dma polymers, the weight average molecular weight values from the solvent system are lower than the values obtained using N a C l (Table III). Except for the very low molecular weight sample, E , the reproducibility o f the epi/dma polymers was fairly good (+/- 10%) from run to run, and from day to day. The dadmac polymer data showed greater variation and concentration dependence. M z and M w increased with increasing concentration, while M n decreased. The solvent system for this particular polymer may not be optimized. Future work will include the use o f different solvent systems to obtain more consistent, reproducible results.

L1NO3

Intrinsic Viscosities. Figure 10 shows the correlation between intrinsic viscosity and weight average molecular weight. A greater increase in intrinsic viscosity is seen with increasing molecular weight for the dadmac polymers, compared to the epi/dma polymers. This Figure also shows the trend toward higher molecular weight values measured in 1 M N a C l (solid symbols) compared to 0.2 M L i N O 3 / 0 . 1 % T F A (hollow symbols). For both the epi/dma and dadmac polymers, the log o f molecular weight is nearly linear with the log o f intrinsic viscosity, but the data show some curvature. This curvature may be due to changes in the polymer conformation with increasing molecular weight. The linear correlation coefficients and the M a r k Houwink constants are summarized in Table V . The Mark Houwink constants for the two solvent systems are within experimental error o f each other and o f the theoretical range predicted by Flory (5).


10.

MEYER & VASCONCELLOS

Molecular-Weight

Distributions

< H Ν©


6

S

Ο

Ο Ο Ο

οο ο ο

W

(Ν PH

οΟ

Ο

ON m

Η Ο Ο

IP

3 Ο

S

ο οο ο —·

η οο

οο ο —· ^ οο

CO

Ο

Ο

Ο

Ο Ο

οο ο ο

W

S'il

Ο m

Ο Λ ζΛ

OH

~

C/3

Ο

ο

m m

υ x

'S

SS

Ή »

a i>

ON Ο CN ι— ON ON O Ο Ο 0

1

f ^



5.00E+00 τ O.OOE+00 + -5.00E+00


-1.00E+01 -1.50E+01 -2.00E+01 ο

til

+ ο

^-

Ι-

oo

Ο

ο

+ ο

ο m +

en

to

Ν) ON

m

+ ο

+ ο

m

+ ο

Elution Volume (mL) Figure 1. 1% A

Figure 2. 1 % B


Molecular- Weight Distributions


τ 1.60E+02

3.50E+01 3.00E+01

-RI LALLS

2.50E+01 2.00E+01 1.50E+01 1.00E+01


5.00E+00 0 . 0 0 E + 0 0 ·• -5.00E+00

-

-1.00E+01 -1.50E+01

4

6.00E+01

\i

-2.00E+01

4.00E+01

U\ σ\

to

-J

m

m

ο

+ ο ο

m

+ ο

·—

!°

NO

-J

m

+ ο

+ ο

Elution Volume (mL) Figure 3. 2% C

3.10E+02

60E+02

-RI

40E+02

c οD -

3.00E+02

LALLS

20E+02

2.90E+02

00E+02 2.80E+02

00E+01 +

00E+01 00E+01 00E+01 00E+00

+ +ν

v

Î

1

00E+01 I I I I I I

00E+01

ô

m +

I I M

+

+

m +

>

r

£

2.70E+02 U 2.60E+02

g

2.50E+02

ο

2.40E+02

§

2.30E+02 2.20E+02

ON

+ ο

Elution Volume (mL) Figure 4. 4% D


138


80E+02

2.10E+02

60E+02 %

c ο

Q. C O

ON Ο

m +

ο ο

bo

ON

m +

ο ο

+

ο

+

ο

to

to

2.00E+02 ο 1.00E+02 " O.OOE+00

to ON

to

m +

ο

Ο

^-

+

Elution Volume (mL) Figure 6. 1.5% W


10.


Molecular-Weight

6.00E+01 τ

3.50E+02 -RI

5.00E+01

3.00E+02

LALLS

r

a 4.00E+01 ο αΙ 3.00E+01 »

2.50E+02

r 00

-ρ 2 . 0 0 E + 0 2

Ο

§

139

Distributions

1.50E+02

2.00E+01 -

g

οο

g


5

S

û S

1.00E+02 §

1.OOE+01 +

vu

O.OOE+00

I I I I I I M I I I I I I I I II I I I I I I I

-1.OOE+01

ô

ο

to

ο ο

ο ο

m +

m +

m + ο

H-

tO

+

+

ο

ο

5.00E+01 O.OOE+00

Elution Volume (mL) Figure 7. 1 % X

τ 2.50E+02

8.00E+01 τ 7.00E+01

+ 2.00E+02

6.00E+01

C/5 ω OS*

5.00E+01 + 1.50E+02

4.00E+01

— I 3.00E+01 Ο

ο Û

2

+ 1.00E+02

2.00E+01 1.OOE+01

+ 5.00E+01 &

O.OOE+00 Ρ O.OOE+00

-1.OOE+01

Elution Volume (mL) Figure 8. 1.5% Y




140

4.00E+01 3.50E+01 3.00E+01 2.50E+01 2.00E+01 1.50E+01 1.OOE+01 5.00E+00 O.OOE+00 -5.00E+00 -1.00E+01 -1.50E+01

4.20E+01 4.00E+01

Elution Volume (mL) F i g u r e d 3.5% Ζ Table V. Least Squares Analysis of Log |η] vs Log Mw

Polymer

a Log (K) corr.coeff.

Static LALLS 1 M NaCl EPI/DMA

GPC/LALLS 0 2MLiNO EPI/DMA

Static LALLS 1 M NaCl DADMAC

GPC/LALLS 0.2 M LiN0 DADMAC

0.459 (+/-0.093) -3.06 (+/-0.42) 0.94

0.434 (+/-0.082) -2.87 (+/-0.35) 0.95

0.521 (+/-0.055) -3.01(+/-0.29) 0.99

0.548 (+/-0.063) -3.02 (+/-0.32) 0.99

3

3



10.


w

ι

ι

3.0

3.5

Molecular-Weight

ι

4.0

I

4.5

141

Distributions

I

5.0

I

5.5

Log Mw S t a t i c LALLS EPI/DMA Polymers GPC/LALLS EPI/DMA Polymers S t a t i c LALLS DADMAC Polymers GPC/LALLS DADMAC Polymers y=x ( 0 . 4 5 9 ) - 3 . 0 6 S t a t i c EPI y=x (0 . 4 3 4 ) - 2 . 8 7 GPC EPI y«x ( 0 . 5 2 1 ) - 3 . 0 1 S t a t i c DADMAC y=x (0 .548) - 3 . 0 2 GPC EPI Figure 10. Intrinsic viscosity as a function of molecular weight, E P I / D M A and D A D M A C polymers


I

6.0

142

M A C R O - I O N

C H A R A C T E R I Z A T I O N

Discussion


For static L A L L S , 1.0 M N a C l was chosen as the solvent for two reasons: consistency with historical data generated on intrinsic viscosity measurements, and to minimize polymer extension. It is well established that in the presence o f sufficient electrolyte polyelectrolytes will behave as neutral macromolecules. The solvent system chosen for G P C / L A L L S was based on two criteria as well: the elimination o f ion exclusion effects and the minimization o f column corrosive effects caused by the presence o f chloride ions (6). The static L A L L S M w values measured in 1 M N a C l are higher than those generated by static L A L L S in 0.2 M L1NO3/TFA or G P C / L A L L S in 0.2 M L1NO3/TFA (the latter two are in agreement), but comparison o f the 'a' constants for the M a r k Houwink equation do not indicate any coil expansion as might be expected. A possible explanation may involve comparison o f both the concentration and the nature o f the solvent system. When a salt is added to a polyelectrolyte solution, the counterions in the coils screen the cationic sites and the electrostatic interactions cause a reduction o f the hydrodynamic dimension o f the polymer. Based on these effects, the more highly concentrated 1 M N a C l solution is expected to exhibit a lower M w than that measured in 0.2 M L1NO3. It has also been shown by Boussouira et al. (7) that for cationic ammonium copolymers, increased shielding o f the polyion occurs when NO3" is used relative to C f . This behavior was in agreement with the affinity order for an ammonium type anion exchanger. The increased shielding would result in a more tightly coiled polymer, which would manifest itself as lower M w . Coupled with the effects the T F A and lower p H , the increased shielding ability o f NO3" may account for the lower M w measured in this solvent system. The competing effects o f the concentration and the nature o f the solvent system may then explain why the M a r k Houwink constants for these two solvent systems are nearly identical. The influence o f the solvent choice on the M w will be explored in a future publication.

Conclusions The technique o f G P C / L A L L S can be used to measure the molecular weight distribution o f water soluble cationic polyelectrolytes typically used for industrial water treatment processes. Weight average molecular weights calculated from the G P C elution curves for epi/dma and dadmac polymers in 0.2 M LiNO3/0.1% T F A are consistently lower than those measured using 1 M N a C l . The different salts used in these two experiments are postulated to contribute to this effect.

Acknowledgments The authors would like to acknowledge Dean S. Bauer o f the Betz Chromatography Services Group for developing the buffer system used in these experiments, and thank him for his invaluable assistance throughout this project.


10.


Molecular-Weight Distributions

143

Literature Cited (1) (2)

(3) (4) (5)


(6) (7)

Betz Handbook of Industrial Water Conditioning; Moore, B. C., et al., Eds.; Ninth Edition; Betz Laboratories, Inc.: Trevose, PA, 1992; pp 24-25. Nagy, D.J.;Terwillinger, D. Α.; Lawrey, B. D.; Tiedge, W. F. In International GPC Symposium 1989 Proceedings; Waters/Millipore Corporation: Milford, M A , 1989; pp 637-662. Potschka, M . J. Chromatography 1988, 441, pp 239-260. Herold, M . American Laboratory 1993, March, pp 35-38. Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1953; pp. 280, 606. Gooding, D. L.; Schmuck M . N.; Gooding, K. M . J. Liq. Chromatogr. 1982, 5, 2259. Boussouira, B.; Ricard, Α.; Audebert, R. J. Poly. Sci.: Part Β 1988, 26, 649-661.

RECEIVED August 6, 1993


Macro-ion Characterization - ACS Publications - American Chemical

Recommend Documents