L_• :E:

,=C

HIGH EFFICIENCY AQUEOUSGEL PERMEATIONCh u_..j J:

,.i 0

By Rick Niel son _ _ "_-._.

Mr. Nielson is Polymer ApplicationsChemist, Industrial o

Waters ChromatographyDivision, MiIlipore Corporation.

The new line of Waters UltrahydrogelTM high-efficiencyaqueous gel

permeationchromatography(GPC) columnswill allow an analyst to perform size

separationsof water-solublepolymers ranging in molecularweight from a few

hundredto severalmillion.

These new columns are packed with a hydroxylatedpolymethacrylatebased

gel, with pore sizes ranging from 120 Angstroms (for the Ultrahydrogel120

column)to 2000 Angstroms (for the Ultrahydrogel2000 column). The

correspondingchain length exclusionlimits, as determinedwith poly(ethylene

oxide) (PEO) standards,vary from 5 x lO3 to an estimated 2 x lO7. Table

l summarizessome of the importantcharacteristicsof these columns.

These columns offer many advantagesover conventionalaqueous GPC columns,

such as a wide pH range (2-12),compatabilitywith high organic aqueous eluent

concentration(up to 20% organic; 50% organic if introducedby gradient),or

greatermobile phase flexibility,and minimal non-sizeexclusioneffects. In

addition,these columns are considerablymore efficient (much narrower peak

widths and higher plate counts)than conventionalaqueous GPC columns.

-455-

MATERIALSAND METHODS:,/

Each column was evaluated using poly(ethyleneoxide) (PEO) standardsfrom

18,000 to 996,000 molecularweight (availablefrom Waters), polysaccharide

(pullulan)standardsfrom 5,800 to 853,000 molecular weight, and poly(ethylene

glycol) (PEG) standards,from 440 to 12,600 molecularweight.

The eluents used for this work were distilled water (phosphatebuffered to

pH 7.0) and O.IM sodium nitrate, with column temperaturesof 30° and 45" being

evaluated. The two eluents could be used interchangeablyfor non-ionic

polymers. The O.IM sodium nitrate should be used for some of the anionic

polymers.

The flow rate in most cases was 0.8 ml/min. The Waters 590 Programmable

Solvent Delivery Module was used. Detectionwas accomplishedwith the Waters

410 DifferentialRefractometer. Data reduction was carried out by the Waters

840 Data and ChromatographyControl Stationwith Waters ExpertTM GPC

software.

SAMPLE/STANDARDSPREPARATION:

In addition to the poly(ethyleneoxide), polysaccharideand poly(ethylene

glycol) standards,samplesof dextrans, gelatin, agar, poly(vinylalcohol),

carrageenans,hyaluronicacid, sugars and polyacrylamidewere

chromatographed. The concentrations(w/v) of the standardsand samplesvaried

from 0.02% to 0.10%, dependingon the molecular weight. All solutionswere

filtered through a 0.45 micron MilliporeMillexR-Hv Filter prior to

injection.

APPLICATIONS:

The Ultrahydrogelcolumns offer excellent resolution for such variedwater

soluble polymersas methyl-cellulose,poly(vinyl alcohol), polyacrylamide,

polyvinylpyrrolidone,in addition to the poly(ethyleneoxide), poly(ethylene

glycol)and polysaccharidediscussed previously. The Ultrahydrogelcolumns

-456-

have also been successfullyused to determinethe molecularweight

distributionsof anionic polymers such as alginic acid sodium salt,

polyacrylicacid sodium salt and sodium polystyrenesulfonate,and also

cationic polymers such as glycol chitosan,DEAE dextran and

poly(N-methyl-2-vinylpyridinium) iodide salt.

The methacrylate-basedgel packing in these columns has a slight negative

charge due to a small amount of residual carboxyl groups. Therefore,in

analyzinganionicor cationic polymers,the chromatographerhas to be

concernedwith ionic effects (such as ion exchange, inclusion,exclusion,

etc.). Dependingon the ionic nature of the polymer (andwhether the polymer

is hydrophilicor hydrophobic),the mobile phase has to be carefullychosen to

minimize non-size exclusioneffects. Adding acetonitrileat a 20% level to

O.IM sodium nitrate, for example, will allow the successfulseparationof

anionic and non-ionichydrophobicpolymers.

RESULTS AND DISCUSSION:

All the poly(ethyleneoxide), poly(ethyleneglycol) and polysaccharide

standardswere chromatographedat the 30°C and 45°C temperaturesusing the

phosphate-bufferedwater (pH = 7) mobile phase. The standardswere also

chromatographedat 45°C using the O.IM sodium nitratemobile phase. The

poly(ethyleneoxide) and poly(ethyleneglycol) standardsfit the same

calibrationcurve for all of the columns tested (See Figure l). The

polysaccharidestandard curves (Figure2) shifted to a slightlyhigher elution

volume, indicatinga smalleroverall molecularsize. This was not observed

for the Ultrahydrogel250 column,which has the lowest pore size of the

columns tested. Note the excellent linearityof the two UltrahydrogelLinear

column curves. The UltrahydrogelLinear and 2000 column calibrationcurves

are linear up to the highest standard (just under l million). Although the

minimum plate count for the UltrahydrogelLinear column is 7,000 plates, plate

counts between I0,000 and II,500 p/ft were obtained for the columns that were

evaluated. Using three UltrahydrogelLinears in series at 45°C will provide

an analyst with tremendousefficiency (over 30,000 theoreticalplates).

-457-

Figure 3 shows the chromatograms(two overlays) for poly(ethyleneglycol)

standards,separating the molecular weights from II,250 down to 440 in

approximately13 min. Three poly(ethyleneoxide) standards,594,000,86,000

and 39,000,were separated in less than 10 min. on an Ultrahydrogel500

column, as shown in Figure 4. The plate count for this column averaged ll,o00

p/ft. Figure 5 illustratestwo different injectionsof poly(ethyleneoxide)

standards. The column used in this case was a single Ultrahydrogel1000, and

the chromatogramshows separationof molecular weights ranging from 18,000 to

996,000. Three polysaccharidestandardswith molecularweight of 23,700,

186,000 and 853,000 separatedon an Ultrahydrogel2000 column are shown in

Figure 6. Notice that in all cases the chromatogramsfor these standardsshow

peak shapes that are narrow and symmetrical.

The last standard chromatogram(Figure7) consistsof _n overlay of two

separate injections of polyethyleneoxides. The column used is a single

UltrahydrogelLinear column using O.IM NaNO3. It was possible virtuallytosuperimposethe poly(ethyleneoxide)/poly(ethylenegl_col) calibrationcurve

on the polysaccharidecurve with the O.IM NaNO3. There is an approximate0.3 mL retentiondifference in the phosphatebuffer mobile phase.

Itwas decided to use the O.IM NaNO3 mobile phase for the majority ofaqueous polymers,especiallywhen some non-size exclusion (ionic)effectsare

suspected to be likely to occur. Figure 8 shows separate and overlaid

chromatogramsof broad distributiondextran standards. These were separated

on two UltrahydrogelLinear columnsand had weight-averagemolecularweights

(Mw, determined by light scattering)of 10,000, 42,000 and 71,000. From the

PEO/PEG calibrationcurve, Mw values of 15,000, 47,000 and 71,000,

respectively,were obtained. The dispersivitieswere approximately1.5 for

all three dextrans.

For a polyacrylamide(See Figure 9), which was said to have a viscosity

averagemolecularweight of 4 million, the author obtained a peak molecular

weight of just under 3 million. The accuracy of this value is somewhat

questionable,since the highest standardwas 996,000 molecularweight and the

calibrationcurve was extrapolatedto higher molecularweights. Because the

concentrationof this sample was only 0.03% (w/v) due to the high molecular

weight, there is a slight increase in the baseline noise on the chromatogram.

-458-

Figure lO illustratesthe GPC chromatogramof a broad MWD poly(vinyl

alcohol) (PVA) sample. This sample was said to have been "pure"and "fairly

monodisperse,"but one can readily observe that there is a significantlow-end

tail, as well as a componenteluting at MW,-1200. The peak molecularweight

for this poly(vinylalcohol) was 60,000.

The chromatogramsof two different samplesof agar are shown in Figure

If. They were from different sources, but thought to be of the same molecular

weight. The overlay clearly shows a difference in the molecularweight

distributions.

The next chromatogram(Figure12) is that of a gelatin sample,separated

on a set of two Linearsplus one 250 Ultrahydrogelcolumn. The distribution

appears to be almost bimodal,with a shoulder being observed on the high

molecularweight end of the curve.

Figure 13 illustratesthe GPC chromatogramsof three different

carrageenans. Carrageenansare used extensivelyin the food industry as

thickeners,or gelling compounds. These samples are different not only in the

MWD, but also in the presenceof a low molecularweight component for the #1

sample. One would expect these three samples to have markedly different

gelIing characteristics.

The next chromatogram(Figure14) is that of a sample of hyaluronicacid.

Hyaluronic acid is used extensivelyin ophthalmic surgery,and also in

treatmentof inflammatoryand degenerativebone diseases. In most cases the

higher the molecularweight, the more positive the therapeuticresponse. We

were able to correlatethe molecularweights of hyaluronicacid sampleswith

their respective intrinsicviscosities. The GPC procedure,however, has much

better reproducibilityand can be done much faster.

The last chromatograms(Figure15) are those of simple sugars. They were

separatedon two DP columns in*20 minutes. A single DP column can be used to

separate the mono, di, and trisaccharidesin less than lO minutes.

"459-

v

CONCLUSION:

The Ultrahydrogelcolumnsafford highly efficientseparationsof water

solublepolymers by GPC. For hydrophilicpolymers (anionicor non-ionic),

O.IM sodium nitrate solution is recommendedas the eluent. For some anionic

and non-ionichydrophobicpolymers the additionof up to 20% (by volume)

acetonitrilewill prevent non-size exclusioneffects. Cationic hydrophilic

polymers (such as DEAE-dextranand glycol chitosan)requirea mobile phaseof

O.8M sodium nitrate in order to prevent extra column effects. Cationic

hydrophobicpolymers are the most difficultto chromatograph,requiringa

mobile phase of O.SM acetic acid plus O.3M sodium sulfate. As long as an

analystunderstandsthe chemistryof the polymer, choosing the correcteluent

and obtaininggood results should be easy. Minimumcolumn efficiency

specificationsare conservative,and most columns exceed the minimum by a

significantmargin.

There are numerous aqueous polymers in the industrytoday which can be

separatedand characterizedon these columns. The author has mentioned justa

few for the purposeof demonstratingthe mobile phase chemistriesneeded to

characterizeanionic, cationic and neutral aqueous solublepolymers.

Additionalmethods developmentprojects currently are in progress.

-a60-

TABLE 1

Column Pore Minimum Exclusion

Size Efficiency Limit

(A) (plates/col) (PEO)

Ultrahydroge] 120 14,000 5,000

Ultrahydrogel250 250 14,000 80,000

Ultrahydrogel500 500 lO,O00 400,000

UltrahydrogellO00 lO00 l0,000 l,000,000

Ultrahydrogel2000 2000 7,000 20,000,000

UltrahydrogelLinear Blend 7,000 20,000,000

-461-

FIGURE I

PEO * PEG

STANDARDS

500 ¢eI000 _LINF_AR

,oo_ I \'_.,,

¢, o,,

•\ -. \

IOK 1_ ®

\

\ \

6 8 I0 12 14, 16MINUTES

-462-

FIGURE 2

POLYSACC HARIDE

STANDARDS

500 _00 INEARIM

lOOK

%

IOK \

R,\ \1000

25O

IK

6 8 I0 12 14 I$

-463-

F IGURE 3

POLY(ETHYLE NE GLYCOLS)

1,580/_I

I I

ELUENT: HzO-pH:7 ii,z5o j ,FLOW RATE" ' , '

'I 4.,820 j0.8 m_/min

i ICOLUMN 250 , 630i I

J L I

TEMP'45c ' ,J I 440

II

I II

II |

! !

I I

s I

I i

Ii I I i I I I ' I

7 8 9 I0 II 12 13 I._

MINUTES

FIGURE 4

POLY(ETHYLENE OXIDES)

ELUENT H20- pH--7

FLOW RATE'O.8m#io594K

COLUMN" 500

eeK TEMP. "45c

I

I

j# ! !

5 I0 15

MINUTES

FIGURE5

POLY_ETHYLENE OXIDES)145K 86K r_

i i /_39K594K/,^ i _ i

,', , , ,, ELUENT:H20_pH=7I L I I I t

I t " _ f II

' ' ' ,' ' FLOW RATE'O.Smk996K I ' l _ ,

!

' COLUMN" I000t I iI

! I I

o, ' ' leK TEMP:45co_ | I "I I , I

I iII I

I|

t t

f

7 8 9 I0 II 12 13 I_MINUTES

F IGURE 6

POLYSACCHARIDES

IBBK

E LUENT" H20 pH- 7

853K FLOW RATE" 0.8 ml/min

COLUMN" 20002 3.7K

i TEMR:45c,,,.II

t- �4

5 I0 15

MINUTES

FIGURE 7

POLY('ETHYLENEOXIDES)

ELUENT. O.IMNAN03

FLOW RATE: Q8 rnyrnin

COLUMN:LINEAR39K IBK

594K /q

TEMR'45c //_ ,_SK , ,I I

' / i:'(3h I I(3o II _ I

I I

I # I

i t I

8 9 JoII

12MINuTEs

FIGURE 8

DEXTRANS

Mw=71K Mv_IOK

ELUENTQIMNANO 3

FLOW RATE'0.8_io

'-- COLUMNS0

,-, 2 LINEARSI ._1_1

4_ i-,i

'_ tEMP. "45cI

I ! I I I I I

19 20 2t 22 23 24 25

MINUTES

FIGURE 9

POLYACRYLAMIDE

M_,_O00,O00

ELUENT'QIM NANO_

FLOW RATE:0,Sinai,COI--

" COLUMNS' 2 LINEARS0

Hd

I ....I

'-' TEMP. 45c-,4'0

I

0

16 17 18 19 20 21 22 23 24MINUTES

FIGURE I0

POLY (VINYL ALCOHOL)

MW:SO,OO0 ELUENT" 0.1M NANO3

F LOW RATE' 0.Sin#it,

COLUMNS " 2LINEARS

I,-..-J TEMP. "45c0

I1,,-II d

4_ d

I

"-1200

# $ ! # # # # I

t8 t9 20. 2 t 22 23 24 25 26

MINUTES

FIGUREII

AGARS

F_.LUENT: 0.1M NANO 3

FLOW RATE" O.B ml/min

COLUMNS" 2LINEARS

\

\kI

-,-11%)I

I

L i i !

15 20 25

MINUTES

FIGURE 12

GELATIN

ELUENT' O.iM NANO 3

FLOW RATE'.:ImVmin

COLUMN S:2 LI NEARS.i"

I 25O

DETECTION dRlrt'}

.._Io

H_J_.I

,,..]L_I

L O J I | | 0 0 e O O 0 O e I !

t6 t7 t8 t9 20 2i 22 23 24 25 26 27 28 29 30 3i

MINUTES

riG.13CARRAGEENANS

9 I: I.OM Mw ELUF..NT;O.IN NaNO 3

F LOW RATE:O.8ml/t'nin

I( 9: I.EIMIglw COLUMNS: 2 LINEARS

I0: 1.3M Iglw DETECTION" dRIZ0HI.-.,it11:t-Z

I Iii4_ u,,4 Z

=" ° '10! U

LOG MOLECULAR WGT.

FIG. 14HYALURONI C ACI D

lot 3821

Mw--2.,674,0O0

ELUENT: O.IN NAN03 i

FLOW RATE: l.Oml/minICOLUMNS: 2 2000+!

7- I 25OOH

I.----_ DETECTION " RItri--zi,i

i u

•..4 oi..,i

i

LOG [MOLECULAR WEIGHT]