inorganic anions - agilent · 2016. 8. 29. · title: inorganic anions author: willi subject:...

Getting Readyfor GPC-SEC

Analysis

Copyright © 2000 Agilent Technologies

All rights reserved. Reproduction, adaption,or translation without prior written permissionis prohibited, except as allowed under thecopyright laws.

Agilent Part No. G2182-90100

Printed in Germany 3/00

Agilent Technologies Deutschland GmbHChemical Analysis Group EuropeHewlett-Packard-Strasse 876337 WaldbronnGermany


AnalysisStart-up kitsfor analyses with organicand aqueous eluents

Organic GPC-SEC Start-up Kit PN: 5064-8251Aqueous SEC Start-up Kit PN: 5064-8252

Upon receipt please verify that all kit contents listed on page 6 (OrganicGPC-SEC Start-up Kit) and page 30 (Aqueous SEC Start-up Kit) areincluded. If any part is missing, contact your Agilent Technologies salesoffice.

Table of Contents:

Part A Getting Ready for GPC-SEC Analysis with Organic Eluents 5I. Introduction 6II. Kit contents and other supplies 6III. Instrumentation 7IV. Preparation 8V. Instrument parameters 9VI. Data acquisition 11VII. Calibration of the GPC system 13VIII. Interactive calculation of molecular weight data 17IX. Automatic calculation of molecular weight data 21X. Applications 24

Part B Getting Ready for SEC Analysis with Aqueous Eluents 29I. Introduction 30II. Kit contents and other supplies 30III. Instrumentation 31IV. Preparation 32V. Instrument parameters 33VI. Data acquisition 35VII. Calibration of the GPC system 37VIII. Interactive calculation of molecular weight data 41IX. Automatic calculation of molecular weight data 46X. Applications 49

Part C Appendices 55I. Troubleshooting 56II. Literature 58III. Acknowledgement 59

5


Analysiswith OrganicEluents

Organic GPC-SEC Start-up KitPN: 5064-8251

Part A

6

I. Introduction

The kit ”Getting Ready for GPC-SECanalysis with organic eluents” is developedto get you up and running quickly andeasily your polymer samples. It containsone PLgel Mixed-C column (Agilent ordernumber 79911GP-MXC), ready to usemixtures of calibration standards and thismanual for the guided development of thedata acquisition-, GPC data analysis- andreporting part of a broad polystyrene testsample in interactive and automatic mode.To ensure that the kit and the instrumentsare functioning properly, a detailedprocedure and test chromatograms aregiven below. The influences of variousparameters on the precision and accuracyof molecular weight data as well typicalerrors are discussed. An applicationchapter is encluded showing typicalanalyses with PLgel columns.We assume that the user is familiar withthe theory of GPC-SEC, the Agilent 1100Series hardware, the ChemStation and theGPC data analysis software.

Part Quantity Part Number

Column: PLgel Mixed-C, 300 x 7.5 mm, 5 µm 1 79911GP-MXC

Test certificate for the column 1 n.a.

PLgel User Guide 1 n.a.

Mixture of four EasyCal polystyrene standards (vial RED)* 3 n.a.*

Mixture of four EasyCal polystyrene standards (vial BLUE)* 3 n.a.*

Mixture of four EasyCal polystyrene standards ( vial GREEN)* 3 n.a.*

Characterization data for polystyrene standards 1 n.a.

Broad polystyrene test sample(vial Yellow) 3 n.a.*

Instruction manual 1 G2182-90100

The polystyrene standards are available in a 10 x 3 vialquantity with part number 5064-8281.

The following parts/supplies should be ordered separately:


Adapter (only needed with 1100 Series quaternary pump) 1 0100-1847

Tetrahydrofuran (THF) stabilized with BHT (Butylated 2 L To be orderedHydroxytoluene) from local

supplier

(You may also use HPLC grade THF with the drawbacks of a higher baseline drift and ahigher price. For applications with an additional UV detector stabilized THF can be usedwith wavelengths above approximately 250 µm.)

II. Kit contents and other supplies

7

Part A Organic GPC-SEC Start-up Kit

a) Agilent Technologies 1100 Series GPCAnalysis System

including:

• Isocratic pump (G1310A)If alternatively the binary or thequaternary pump is used never mix theeluent with the pump. In case of the binarypump strictly use 100% of channel A and0% of channel B. The quaternary pumpshould be plumbed such that only onechannel is used and the multi channelgradient valve is bypassed. This needs anadditional adapter (P/N 0100-1847) toconnect the solvent tube directly to theactive inlet valve instead of theproportioning valve

• Online degasser (G1323A - stronglyrecommended)

• Autosampler (G1313A - stronglyrecommended)

• Thermostatted column compartment (G1316A- strongly recommended)

• Refractive index detector (RI – G1362A)• ChemStation (G2170AA, revision A.06.04 or

higher) with GPC data analysis software(G2182AA, revision A.01.03 or higher)

For obtaining reliable results theequipment should have passed an OQ/PVtesting before.

b) Column

• Agilent Technologies PLgel Mixed-C,300 x 7.5 mm, 5 µm (79911GP-MXC)

The PLgel Mixed-C column is packed with5 µm particles of different pore sizes andthus enables the separation of polymersover a wide molecular weight range (fromabout 200 to 3 million). Such wide rangesrequire with traditional single pore sizecolumns sets of several columns, typicallybetween 2 to 3. On the other hand a 30 cmlong mixed gel column will never have thesame plate numbers as a complete columnset. Therefore a typical application for thiscolumn is a fast, screening type analysiswhich can be afterwards optimized. Thisoptimization is achieved either through theaddition of further mixed gel or dedicatedsingle pore size columns (refer to chapterX. Applications).

III. Instrumentation

8

IV. Preparation

It is time saving if the chapters IV.Preparation and V. Instrument Parametersare performed in the afternoon of the firstday and the other chapters on thefollowing day. This gives the refractiveindex detector and the column sufficienttime for conditioning during the night.

1. Add 1.0 mL THF to one vial of each of thestandard mixtures (RED, BLUE andGREEN) and one broad sample vial(YELLOW) and allow dissolving for60 minutes. Shake the vials every tenminutes. After addition of the solvent thevials should be used within two days.

2. Switch on all modules of the GPC system

3. Make sure that the system is free of salts orsolvents immiscible with THF. Otherwiseflush the system thoroughly with water oran intermediate solvent

4. Fill 1 liter of stabilized THF into the solventbottle of the 1100 Series isocratic pump(for the binary and the quaternary pumprefer to chapter Instrumentation)

5. Open the purge valve of the pump andpurge the channel in use at 5 mL/min for atleast 10 minutes

6. Close the purge valve and flush the pump,autosampler and column thermostat at1 mL/min for at least 5 minutes withoutcolumn.

9


V. Instrument parameters

1. Analytical conditions:If the GPC data analysis software revisionis A.02.01 or higher you can load thefollowing methods from your hard discusing the below method names and thepath C:\hpchem\X\Methods (where X is theinstrument number):

• OK_STAN.M – for analysis of thepolystyrene standards

• OK_SAM.M – for analysis of the broadpolystyrene test sample

The two methods differ in the integrationparameters and in the ”Run TimeChecklist”. For the narrow standards andthe broad polystyrene sample differentparameters are needed because ofdifferent peak shape. They may needspecial adaptation to the requirements ofyour peaks determined by the GPC systemand column performance (refer to chapterIX, Automatic calculation of molecularweight data).The analysis of polymer standards is aconventional HPLC run and does notrequire the activation of the GPC dataanalysis software in the ”Run TimeChecklist”.

If your GPC data analysis software has arevision number lower than A.02.01 key inthe following parameters and save themafterwards using the above path and filenames:

RemarksPump:Eluent channel: 100%A strictly isocratic, no premixing!

Refer to chapter III. InstrumentationFlow rate: 1.0 mL/minStop time: 12 minMax. pressure limit: 150 barMin. pressure limit: 5 bar

Autosampler:Injection volume: 50 µL standards and test sample

Column thermostat:Temperature: 25 °C left and right side

Refractive index detector:Optical unit temperature: 25 °CPolarity: PositiveAutomatic recyclingafter analysis: OFFPeak width: 0.1 min

Use the enhanced integrator (both standard and test sample analysis).

10

Integrator settings for broad polystyrenetest vial (method OK_SAM.M)(will need special adaptation to peak shapeand retention times with your system):

Integrator settings for polystyrenestandards(method AK_STAN.M):

RemarksInitial Slope Sensitivity: 0.500Initial Peak width: 0.500Initial Area Reject: 10000Initial Height Reject: 1000Initial Shoulders OFF

Time 0.000 Negative Peaks ON Negative peaks are integrated wellTime 0.00 Integrator OFFTime 3.5 Integrator ON No peak elution possible before

3.5 minutes

Remarks

Initial Slope Sensitivity: 1.00 Adapted to broad polymer peakInitial Peak width: 1.00 Adapted to broad polymer peakInitial Area Reject: 10000Initial Height Reject: 1000Initial Shoulders OFF

Time 0.000 Negative Peaks ON Negative peaks are integrated wellTime 0.000 Integration OFF No elution of peaks possible before

volume of total penetrationTime 4.00 Integration ONTime 4.010 Baseline NOW Forces integrator to reset baseline

2. Change the flow rate to 0.5 mL/min

3. Install the column in the thermostattedcolumn compartment and connect theinlet of the column to the autosampler andthe outlet to the refractive index detector.Pump eluent at 0.5 mL/min to waste for 30minutes to replace the shipping solvent(toluene) with the eluent (THF). Duringthis time purge the reference cell of the RIdetector.

4. Increase the flow rate to 1 mL/min andpump for at least another 2 hours, at leastuntil the detector baseline is stable, nodrift and noise can be seen at a full scalesetting of 5000 nRIU(detector noise: £ ± 1 x 10–8 RIU,detector drift: £ 3 x 10–6 RIU/h,wander: £ 3 x 10–7 RIU). For the automatedcalculation of these detector parametersrefer to the Verification (OQ/PV) part ofyour ChemStation.

5. For best precision and accuracy it isrecommended to perform thisconditioning over night.

11

VI. Data acquisition

1. Set up a sequence of two 50 µl injections ofeach standard (method AK_STAN.M) andthe broad test vial (method OK_SAM.M) inthe order of PS vial RED, PS vial BLUE, PSvial GREEN and broad PS vial YELLOW.

2. Check for baseline stability and start thesequence

3. Typical chromatograms for the injectionswith HPLC grade THF as mobile phase andas sample solvent are shown below.


Figures 1 – 3:Typical chromatograms of calibration mixtures PS vialRED, PS Vial BLUE and PS vial GREEN

12

The first 4 peaks of the standardchromatograms are the narrowpolystyrene standards with decreasingmolecular weight. After the last standardpeak elute monomer peaks due to solventimpurities. If you use stabilized THF youwill see a large negative peak of thestabilizer. The other negative peaks aremainly caused by pressure drops duringthe injection cycle.These monomer and pressure drop peakscan vary from instrument to instrumentand with mobile phase and sample solvent.

Figure 4:Typical chromatogram of the polystyrene test sample(YELLOW)

13

VII. Calibration of the GPC system

Figure 5:Typical overlay of two consecutive calibrationruns from vial BLUE

If the standard and the test samplechromatograms fulfill these conditions usethem for calibration and evaluation, if notrefer to chapter Troubleshooting. Becauseof this good precision of the retentiontimes use for the present calibration onlythe second run of each vial. For a “reallive” calibration and application you mayuse all data points.

1. In ChemStation Data Analysis loadOK_STAN.M

2. In the [GPC] menu bar select [ActivateGPC]

3. In [GPC Settings ...] inspect whether theGPC settings reflect your hardwareconfiguration (e.g. detectorconfiguration). Make sure that thedefault.cal calibration file is loaded and

Inspect the chromatograms carefully forrepeatability of peak shapes and retentiontimes to avoid stray points. With a goodperforming GPC system the retentiontimes of two consecutive injections of thesame polymer should in a maximum differin the positions two and/or three behindthe point. Consecutive injections from thesame vial should look like a singlechromatograms when overlaid. Thisshould be the case for the positive peaks,the negative peaks may have a varyingheight caused by the pressure drop duringthe injection cycle. For a typical examplerefer to figure 5.


14

that the “Report Settings” section of theGPC-Settings dialog are set to “InteractiveScreen Review”. Press the OK button whenfinished to return to the Data Analysis viewof the ChemStation.

4. Load the second data file of polystyrenestandard vial RED and transfer it to theGPC data analysis software by pressing“Calculate GPC Results” of the GPC menu.

5. Highlight the Raw Data window (R) andopen the sample editor by selecting[Sample] from the [Editor] menu. Enterthe molar masses Mp of the 4 narrowstandards of polystyrene standard vialRED with decreasing molar mass as theyare displayed in the Certificate of Analysis.All the other inputs are not essential forthe narrow standard calibration. Press theOK button when finished. Select [Window |Calibration] from the GPC data analysismenu and create an empty calibration fileby clicking on [File | New]. The calibrationwindow will change background color.

6. Highlight the Elugram window and movethe cursor below the (first) peak andbelow the x-axis. Click on the right mousebutton and select [Find Maximum] fromthe pop-up command box. There might bea difference in the retention timescalculated by the ChemStation dataanalysis and the GPC data analysis partdue to different algorithm used. In the Addto calibration dialog box select the propermolar mass of the calibration standardfrom the list by clicking on the correctradio button. Then add this calibrationpoint to the calibration table by pressingthe “add to calibration” button. Continuethis for all standards in this chromatogram.

7. Then load the data files of polystyrenestandard vials BLUE and GREEN from theData Analysis and process them in thesame way as described in steps 4 – 6 untilyou have added all calibration standards tothe calibration table.

8. When you have completed the calibrationtable, activate the Calibration window inthe GPC ADD-ON window and choose aregression model, e.g. Polynom 3, fromthe “Fit” drop-down selection list in orderto create a calibration curve.

15

9. Try all possible regression models andoptimize with the deviation data displayedin the calibration table. The fit qualityshould in addition always be controlled bythe signs of the differences between theoriginal calibration points and the valuescalculated by regression, using theabsolute M-values and not their logarithms.The signs of these residues must berandomly distributed. The fit is insufficientif this is not the case (1).

10. You can check quickly the quality of theregression with the deviation plot to beselected from the “Compare” drop-downselection list.Figure 6 shows a typical calibration curvewith a 3rd order polynom fit. The pointsshowing deviation are mostly locatedbelow the calibration curve indicating aninsufficient fit.Figure 7 shows the same calibration datawith a 7th order fit. This fit is superior asshown by the random distribution of thedeviation points and the lower deviations.

Figure 6:Calibration curve for polystyrene standards obtainedwith an insufficient 3rd order fit (deviation points notrandomly distributed)


To rely only on the above requirementscan easily create errors. For example, for 6calibration points it is always possible to fita polynomonal of the fifth degree suchthat the calculated curve runs through all

Figure 7:Calibration curve for polystyrene standards obtainedwith a 7th order fit (deviation points randomlydistributed)

16

calibration points. However you usuallyreceive swinging curves, i.e. curves whichhave partially increasing slope, which isphysically meaningless. The followingrequirements should be fulfilled inaddition:

• The slope of the curve should bephysically meaningful. You can view thederivative of the calibration curve in therespective column of the editor sectionand as graphical information by selectionfrom the Compare list

• The slope of the calibration curveshould be highly negative for small andlarge elution volumes while there shouldbe a broad region with relativelyconstant value within.

11. Choose [file | print] to make a printout ofthe calibration table and graph.

12. Save the calibration file using the [Save As]dialog from the File menu and give adescriptive name for this calibration file,e.g. OK_Calib.cal.

17

VIII. Interactive calculation of molecular weight data

1. In ChemStation Data Analysis load themethod OK_SAM.M.

2. Load a data file obtained for the broadpolystyrene test sample.

3. Select [GPC], then [GPC Settings] andload the calibration file OK_Calib.cal youhave saved in the end of chaptercalibration

4. Press OK and then Save methodOK_SAM.M, the calibration fileOK_Calib.cal is now saved with the methodOK_SAM.M

5. Next select the [GPC] menu and then[Calculate GPC Results]

6. The file is transferred to the GPC dataanalysis software which displays the GPC-Addon top-level window with the broadtest sample as example

Fig. 8:Top-level window of the GPC data analysis softwarefor broad PS test sample

a) Defining the baseline

7. In the raw data-(R) part define next thebaseline by moving the red triangles to thecorrect positions. Ideally the baselineshould be taken as a straight line betweenthe elution prior to the size exclusion limit,about 4 mL, and that after the last impuritypeak, i.e. areas in which no elution cantake place in an ideal GPC separation(1,3),refer to figure 9. For setting the baselinecorrectly make sure by zooming in toabout 10% of total peak height that there is

• no peak elution at the position of thered triangles

• no cut off of the polymer peak


18

If this drawing of the baseline does notwork properly – e.g. peaks from previousanalysis elute prior to the size exclusionlimit or the baseline is not sufficientlystable after the last impurity peak – set thebaseline markers exactly in the positionwhere the polymer peak starts and ends.For better seeing this you have to zoom in.

Baseline setting as shown in figure 10 hasthe drawback of less flexibility for settingthe integration limits in the Elugram (E)window (refer to below) but is often amore practical approach.

Figure 9:Raw data window for broad PS test sample with baselinemarkers set prior to volume of total exclusion and afterlast impurity peak

Fig. 10:Raw data window for broad PS test sample with baselinemarkers set at peak start and end

baselinemarker

baselinemarker

19

b) Defining the integration limits

In the Elugram window the baselinecorrected raw data are presented.In the Elugram window the integrationlimits for calculation of the molecularweight distribution and the molecularweight averages have to be set with the redmarkers. Position them exactly before thestart and exactly at the end of the peak.You will have to zoom in for better seeingthis. This step is typically not needed if youset the baseline as shown in figure 10.

During the optimization of baseline andintegration limits study the drasticinfluence of the position of the triangles onthe molecular weight results. You canchange especially the Mn-value drasticallywith the set points. This strong influence ofthe integration limits and the baselinesetpoints on the molecular weight data isone of the reasons for the limited precisionand reproducibility of the molecularweight averages.

A typical value for the Mw-value of thebroad polystyrene sample under thechosen conditions is:

Fig. 11:Elugram window showing correct positionof integration limits

Molecular weight Typical GPC value Limits (± 20%)

Mw 265000 212000 - 318000

The Mw-data should be in a range of ± 20%within the typical value. Try to comeclosest to the reference value If youcannot succeed refer to Part CAppendices, chapter I. Troubleshooting.


20

8. Highlight the [Mass Distribution(M)]window, then select the [Raw Data] menuand press [Print]. You will get a single pagereport printed consisting of the massdistribution, the main method parametersand in the bottom the molecular weightdata.

9. The molar mass distribution should looklike the one shown in figure 12.

Fig. 12:Typical molecular weight distribution of the broadpolystyrene sample

10. Determine the molecular data in the sameway for the second data file obtained withthe broad polystyrene vial (YELLOW).

Hardware Parameters Software Parameters

• Column performance • Accuracy of calculation procedures• Flow precision • Quality of baseline setting• Temperature precision • Quality of setting the calculation• Baseline stability start and end marks• Injection volume precision • Quality of calibration curve• Quality of the standards • Number of data points

• Possibility to use an internal standardcorrection for flow rate changes

The lab to lab reproducibility of molecularweight data depends on several hardwareand software parameters as the tablebelow shows:

Because of the many influencingparameters lab to lab reproducibility isoften poor. R. Bruessau reports (1) of anEuropean round robin test whichresulted in differences from lab to lab forMn of ± 16 % and for Mw of ± 9%. Similardata were reported from a Japanese roundrobin test(2). In both round robbin studiesonly experienced GPC laboratoriesparticipated!

21

This chapter explains how to set up thesystem for completely automatic analysesconsisting of data acquisition, GPC dataanalysis and reporting based on the abovedeveloped method. An Agilent 1100 SeriesGPC analysis system with degasser, pump,autosampler, thermostatted columncompartment and refractive indexdetector is required.

IX. Automatic calculation of molecular weight data

a) Preparation and start

1. Make sure that the system is still runningusing the above conditions and thatmethod OK_SAM.M is loaded. In the [GPCSettings] window the calibration fileOK_CALIB.CAL is activated.

2. Load the last data file obtained for thebroad polystyrene test sample.

3. The integration parameters in OK_SAM.Mneed to be optimized for the peak shapeobtained for the broad polystyrene withyour system. The GPC data analysis part ofthe ChemStation calculates in a sequencethe molecular weight data between thestart mark of the 1st integrated and the endmark of the last integrated peak. It istherefore extremely important that

• Only the polymer peak is integrated(refer to figure 4)

• The start and end integration marks ofthe polymer peak are placed where thepeak starts and ends (zoom in severaltimes for better seeing this, refer also tochapter VIII and figure 10)

• For optimizing the start and endintegration marks use:

Remarks

• Slope sensitivity Big impact on start and end mark(Use values as 0.01 or 1000 for studyingthe influence)

• Peak width Big impact on start and end mark(Use values as 0.01 or 10 for studying the influence)

• Negative peaks On Negative peaks are integrated well andbaseline is not drawn to minimum of negativepeaks

• Integration On/Off For suppression of peaks of no interest

• Baseline Now Forces integrator to reset baseline


22

Fig. 13:Example for broad test sample with correctlyset integration marks

For more information on the integrationparameters and events refer to theChemStation help text.

Change the above parameters until thestart and end marks are optimized. Correctdrawing of the baseline itself within thestandard Data Analysis is not important.The GPC data analysis itself draws thebaseline arbitrarily from integration startto end mark.

4. Go into the [GPC Settings…] of the[GPCmenu] and

• Change the Report Settings from“Interactive Screen Review” to“Print Results”

• Configure a report consisting of e.g.“Graphical MWD”, a “Method”- and a“Results” part

5. Save the method again and set up asequence consisting of 6 injections of thebroad polystyrene test sample


b) Evaluation

1. You should have received the followingprintouts for each run automatically

• One standard HPLC report• One single page GPC report consisting

of one “Graphical MWD”, one “Method”-and one “Results” part

2. In the ChemStation Data analyis overlaythe chromatograms of the six analyses

23

3. The overlay should look with a well-conditioned column and a well-maintainedGPC system visually like a singlechromatogram at least in the polymerregion. If this is not the case the precisionof the molecular weight data may benegatively affected.

The negative peaks may have a varyingheight caused by the pressure drop duringthe injection cycle.

3. Calculate the Standard Deviation (SD) andRelative Standard Deviation (%RSD) forthe 6 obtained Mn- and Mw-values using thefollowing formulas on your pocketcalculator:

Fig. 14:Overlay and zoom in into 6 automatic runsof broad PS test sample

4. The relative standard deviations should be• For Mn < 10 %• For Mw < 7 %

If this is not the case refer to part Cchapter I. Troubleshooting. For moreinformation on precision of molecularweight data achievable with optimized,state of the art equipment refer toliterature 5)


24

X. Applications

a) Example of method development

The chromatogram shows the analyses of abroad technical polystyrene: part a) withone PLgel Mixed-C column and part b)with a set of three PLgel Mixed-C columns.Due to the better separation of the columnset and the reduced influence of bandbroadening, the calculated molecularweight averages are in the later case closerto the values determined with referencemethods, e.g. light scattering (refer totable).

Columns: a) 1 x PLgel Mixed-C, 7.5 x 300 mm, 5 µm(Agilent Part No. 79911GP-MXC)b) 3 x PLgel Mixed-C, 7.5 x 300 mm, 5 µm in series

Mobile phase: Tetrahydrofuran (THF)Column temp.: 25 °CFlow rate: 1 mL/minDetector: Refractive Index DetectorCalibrant: Polystyrene EasyCal standards in vials for calibration

Fig. 15:Broad polystyrene with 1 and 3 PLgel Mixed columns

Reference Mn* Mw** DData 86000 246000 2.86

Difference [%] 9 15 27(1 x PLgel Mixed)

Difference [%] 3.1 4.8 2.0(3 x PLgel Mixed)

* measured by GPC** measured by light scattering

25

b) Example of process control

This figure shows an overlay of 3chromatograms of another technicalpolystyrene – the original granulate andbefore and after injection molding. Afterthe first injection molding, there is almostno change in the chromatogram andtherefore the molecular weightdistribution. After grinding the chips, andinjection molding a second time there is asignificant change which will have an effecton the properties. The visual information issupported by the number averagemolecular weight, Mn, as calculated by theChemStation GPC data analysis software:

Mn (original granulate): 59000Mn (after 2nd process): 55000

Columns: 3 x PLgel Mixed-B columns, 7.5 x 300 mm, 5 µm(Agilent Part No. 79911GP-MXB)

Mobile Phases: Tetrahydrofuran (THF)Flow rate: 1 mL/minOven temp: 20 °CInjection Volume: 10 µLDetector: Refractive Index DetectorSample: dissolved in 1 mL THFCalibrant: EasyCal polystyrene standards in vials were

used for narrow standard calibration

Fig. 16.Example of process control


26

Columns: 3 x PLgel Mixed-C in series, 7.5 x 300 mm, 5 µm(Agilent Part No. 79911GP-MXC)

Mobile phase: Tetrahydrofuran (THF)Flow rate: 1.0 mL/minOven temp.: 20 °CInjection volume: 10 µLDetector: Refractive Index Detector and Diode Array DetectorSample: 26 mg dissolved in 1 mL THFCalibrant: EasyCal Polystyrene standards in vials were

used for narrow standard calibration

Fig. 17:PMMA polymer with Mw = 316000

c) Example of quality control

About 1.5 million tons of PMMA areproduced on a worldwide basis forapplications as

• plastic glass, e.g. for roofs,• safety glasses,• glasses for cars and dishes.

The quality of the polymer stronglydepends on the molecular weight and themolecular weight distribution. The figureshows the overlay of a Refractive Indexand a UV detector signal of a PMMA with aweight average molecular weight (Mw) of316000.

27

d) Example of quality control

Alkyd resins are widely used for theproduction of paints. This example showsthe quality control analysis of two resinsused for high quality paints in the carindustry.

The paint failed in the application whenproduced with the bad quality resin, whileit had excellent adhesive properties whenproduced with the good quality resin.

Fig. 18:GPC-SEC analysis in quality control:Resin quality in paint

Columns: PLgel 10 2 Å, 7.5 x 300 mm, 5 µm (Agilent Part No. 79911GP-501)in series withPLgel 5 x 10 2 Å, 7.5 x 300 mm, 5 µm (Agilent Part No. 79911GP-502)andPLgel 10 4 Å, 7.5 x 300 mm, 5 µm (Agilent Part No. 7991GP-504)

Mobile phase: Tetrahydrofuran (THF)Flow rate: 1.5 mL/minOven temp.: 20 °CInjection volume: 10 µLDetector: Refractive Index Detector

For more application examples refer to“Polymer and Hydrocarbon ProcessingSolutions Guide”, Agilent publicationnumber 12-5968-7020E.


29

Part B Aqueous SEC Start-up Kit

Getting Readyfor SEC Analysis

with AqueousEluents

Aqueous SEC Start-up KitPN: 5064-8252

Part B

30

I. Introduction

The kit “Getting Ready for GPC-SEC withAqueous Eluents” is developed to get youup and running quickly and easily yourpolymer samples. It contains onePL aquagel-OH Mixed column (Agilentorder number 79911GF-MXA, ready to usemixtures of calibration standards and thismanual for the guided development of thedata acquisition-, GPC data analysis andreporting part of a broad dextran testsample in interactive and automatic mode.To ensure that the kit and the instrumentsare functioning properly, a detailedprocedure and test chromatograms aregiven below.The influences of various parameters onthe precision and accuracy of molecularweight data as well typical errors arediscussed. An application chapter isencluded showing typical analyses withPL aquagel columns.We assume that the user is familiar withthe theory of GPC/SEC, the 1100 Serieshardware, the ChemStation and the GPCdata analysis software.


Column: PL aquagel-OH Mixed, 8 mm, 7.5 x 300 mm 1 79911GF-MXA

Test certificate for the column 1 n.a.*

PL aquagel-OH user guide 1 n.a.*

Mixture of four EasyCal polyethyleneoxidestandards PEO/PEG vial RED* 3 n.a.*

Mixture of four EasyCal polyethyleneoxidestandards PEO/PEG vial BLUE* 3 n.a.*

Mixture of four EasyCal polyethyleneoxidestandards PEO/PEG vial GREEN* 3 n.a.*

Characterization data for PEO/PEG standards 1 n.a.

Broad dextran test sample 3 n.a.

Instruction manual 1 G2182-90100

The polyethyleneoxide/polyethyleneglycol standards are available in a 10 x 3 vialquantity with part number 5064-8280.

The following parts/supplies should be ordered separately:


Adapter (only needed with 1100 Series quaternary pump) 1 L 0100-1847

Water (HPLC grade) 2 L To be orderedfrom localsupplier

II. Kit contents and other supplies:

31

III. Instrumentation

a) Agilent Technologies 1100 Series GPCanalysis system

including:

• Isocratic pump (G1310A)If alternatively the binary or thequaternary pump is used never mix theeluent with the pump. In case of the binarypump strictly use 100% of channel A and0% of channel B. The quaternary pumpshould be plumbed such that only onechannel is used and the multi channelgradient valve is bypassed. This needs anadditional adapter (P/N 0100-1847) toconnect the solvent tube directly to theactive inlet valve instead of theproportioning valve

• Online degasser (G1323A – stronglyrecommended)

• Autosampler (G1313A – stronglyrecommended)

• Thermostatted column compartment (G1316A– strongly recommended)

• Refractive index detector (RI – G1362A)• ChemStation (G2170AA, revision A.06.04 or

higher) with GPC data analysis software(G2182AA, revision A.01.03 or higher)

For obtaining reliable results theequipment should have passed an OQ/PVtesting before.

b) Column

• Agilent Technologies PL aquagel-OH Mixed,8 mm, 7.5 x 300 mm, (79911GF-MXA)

The PL aquagel-OH Mixed is packed with 8µm particles of different pore sizes andthus enables the separation of polymersover a wide molecular weight range (fromabout 100 to 10000000). Such wide rangesrequire with traditional single pore sizecolumns sets of several columns, typicallybetween 2 to 3. On the other hand a 30 cmlong mixed gel column will never have thesame plate numbers as a complete columnset. Therefore a typical application for thiscolumn is a fast, screening type analysiswhich can be afterwards optimized. Thisoptimization is achieved either through theaddition of further mixed gel or dedicatedsingle pore size columns (refer to chapterX. Applications).


32

IV. Preparation

It is time saving if the chapters IV.Preparation and V. Instrument Parametersare performed in the afternoon of the firstday and the rest on the following day. Thisgives the refractive index detector and thecolumn sufficient time for conditioningduring the night.

1. Add 1.0 mL water to one vial of each of thestandard mixtures (RED, BLUE andGREEN) and one of the broad sample vials(YELLOW) and allow dissolving for 60minutes. Shake the vials every ten minutes.After addition of the solvent the vialsshould be used within two days.

2. Switch on all modules of the GPC system

3. Make sure that the system is free ofsolvents immiscible with water. Otherwiseflush the system thoroughly with anintermediate solvent

4. Fill 1 liter of HPLC grade water into thesolvent bottle of the 1100 Series isocraticpump (for the binary and the quaternarypump refer to chapter Instrumentation)

5. Open the purge valve of the pump andpurge the channel in use at 5 mL/min for atleast 10 minutes

6. Close the purge valve and flush the pump,autosampler and column thermostat at1 mL/min for at least 5 minutes

33

V. Instrument parameters

1. Analysis conditions:If the GPC data analysis software revisionis A.02.01 or higher load the followingmethods from your harddisc using thefollowing method names and the pathC:\hpchem\X\Methods (where X is theinstrument number):• AK_STAN.M – for analysis of the PEO

standards• AK_SAM.M – for analysis of the broad

dextran test sample

The two methods differ in the integrationparameters and in the “Run TimeChecklist”. For the narrow standards andthe broad dextran sample differentintegration parameters are neededbecause of too different peak shape. Theymay need adaptation to the requirementsof your peaks determined by GPC systemand column performance (refer to chapterIX, Automatic calculation of molecularweight data). The analysis of polymer doesnot require the activation of the GPC dataanalysis software in the “Run TimeChecklist”.

If your GPC data analysis software has arevision number < A.02.01 key in thefollowing parameters and save themafterwards using the above path and filenames:

RemarksPump:Eluent channel: 100% A strictly isocratic, no premixing!

Refer to chapter III. InstrumentalFlow rate: 1.0 mL/minStop time: 12 minMax. pressure limit: 140 barMin. pressure limit: 2 bar

Autosampler:Injection volume: 50 µL

Column thermostat:Temperature: 25 °C left and right side

Refractive index detector:Optical unit temperature: 25 °CPolarity: PositiveAutomatic recyclingafter analysis: OFFPeak width: 0.1 min

Use the enhanced integrator.


34

2. Change the flow rate to 0.5 mL/min

3. Connect the inlet of the column to theautosampler and the outlet to therefractive index detector and pump eluentat 0.5 mL/min to waste for 30 minutes toreplace the shipping solvent (water + 0.1%sodiumazide) with the eluent. During thistime purge the reference cell of the RIdetector.

4. Increase the flow rate to 1 mL/min andpump for another 30 minutes, at least untilthe detector baseline is stable, no drift andnoise can be seen at a full scale setting of5000 nRIU (detector noise:£ ± 1 x 10 –8 RIU, detector drift:£ 3 x 10 –6 RIU/h, wander: £ 3 x 10 -7 RIU).For the automated calculation of thesedetector parameters refer to theVerification (OQ/PV) part of yourChemStation.

5. For best precision and accuracy it isrecommended to perform thisconditioning over night.

Integrator settings for broad dextran testvial (method AK_SAM.M)(may need special adaptation to peakshape and retention times with yoursystem):

RemarksInitial Slope Sensitivity: 0.500Initial Peak width: 0.500Initial Area Reject: 10000Initial Height Reject: 1000Initial Shoulders: OFF

Time 0.000 Negative Peaks ON Negative peaks are integrated well

Integrator settings for PEO standards(method AK_STAN.M):

Remarks

Initial Slope Sensitivity: 1.000 Adapted to broad polymer peakInitial Peak width: 1.00 Adapted to broad polymer peakInitial Area Reject: 10000Initial Height Reject: 1000Initial Shoulders OFF

Time 0.000 Negative Peaks ON Negative peaks are integrated wellTime 0.000 Integration OFF No elution of peaks possible before

volume of total exclusionTime 4.00 Integration ONTime 4.01 Baseline NOW Forces integrator to reset baseline,

e.g. immediately before the peak

35

VI. Data acquisition

1. Set up a sequence of two 50 µL injectionsof each standard (method AK_STAN.M)and the broad test vial (methodAK_SAM.M) in the order of vial RED, vialBLUE, vial GREEN and broad dextran vialYELLOW. The dextran sample inject 3times and discard the first injection.Typically it takes one injection of thedextran until stable conditions for thecolumn are obtained.


3. Typical chromatograms for the injectionsare shown below.

Figures 1 – 3:Typical chromatograms of calibration mixtures vial RED,vial BLUE and vial GREEN


36

Figures 4:Typical chromatogram of the dextran test sample(YELLOW)

37

VII. Calibration of the GPC system

Inspect the chromatograms carefully forrepeatability of peak shapes and retentiontimes to avoid stray points. With a goodperforming GPC/SEC system the retentiontimes of two consecutive injections of thesame polymer should in a maximum differin the positions two and/or three behindthe point. Consecutive injections from thesame vial should look like a singlechromatograms when overlaid. For a goodexample refer to figure 5.

If the standard and the test samplechromatograms fulfill these conditions usethem for calibration and evaluation, if notrefer to chapter Troubleshooting. Becauseof this good precision of the retentiontimes use for the present calibration onlythe second run of each vial. If later on ahigher precision is needed use all datapoints.

1. In Data Analysis load AK_STAN.M

2. In the [GPC] menu select [Activate GPC]

3. In [GPC Settings ...] inspect whether theGPC settings reflect your hardwareconfiguration (e.g. detectorconfiguration). Make sure that thedefault.cal calibration file is loaded andthat the “Report Settings” section of theGPC-Settings dialog are set to “InteractiveScreen Review”. Press the OK button whenfinished to return to the Data Analysis viewof the ChemStation.

Figure 5:Good overlay of two consecutive calibration runsfrom vial Green


38

4. Load the second data file of PE0 standardvial RED and transfer it to the GPC dataanalysis software by pressing “CalculateGPC Results” of the GPC menu.

5. Highlight the Raw Data window (R) andopen the sample editor by selecting[Sample] from the [Editor] menu. Enterthe molar masses Mp of the 4 narrowstandards of PE0 standard vial RED asthey are displayed in the Certificate ofAnalysis. All the other inputs are notessential for the narrow standardcalibration. Press the OK button whenfinished. Select [Window | Calibration]from the GPC data analysis menu andcreate an empty calibration file by clickingon [File | New]. The calibration windowwill change background color.

6. Highlight the Elugram window and movethe cursor below the (first) peak andbelow the x-axis. Click on the right mousebutton and select [Find Maximum] fromthe pop-up command box. There might bea difference in the retention timescalculated by the ChemStation dataanalysis and the GPC data analysis partdue to different algorithm used. In the Addto calibration dialog box select the propermolar mass of the calibration standardfrom the list by clicking on the correctradio button. Then add this calibrationpoint to the calibration table by pressingthe “add to calibration” button. Continuethis for all standards in this chromatogram.

7. Then load the data files of PEO standardvials BLUE and GREEN from the DataAnalysis and process them in the same wayas described in steps 4 – 6 until you haveadded all calibration standards to thecalibration table.

8. When you have completed the calibrationtable, activate the Calibration window inthe GPC ADD-ON window and choose aregression model, e.g. Polynom 3, fromthe “Fit” drop-down selection list in orderto create a calibration curve.

9. Try all possible regression models andoptimize with the deviation data displayedin the calibration table. The fit qualityshould in addition always be controlled bythe signs of the differences between theoriginal calibration points and the valuescalculated by regression, using theabsolute M-values and not their

39

logarithms. The signs of these residuesmust be randomly distributed. The fit isinsufficient if this is not the case (1).

10. You can check quickly the quality of theregression with the deviation plot to beselected from the “Compare” drop-downselection list.Figure 6 shows a typical calibration curvewith a 3rd order polynom fit.Figure 7 shows the same calibration datawith a 7th order fit. This fit is superior asshown by the lower deviation values, inaddition the deviation points are randomlydistributed(figure 8).

Figure 7:Calibration curve for PE0 standards obtained with a 7th

order fit


Figure 6:Calibration curve for PE0 standards obtained with 3rd

order fit

40

Figure 8:Calibration curve for PE0 standards obtained with a7th order fit and randomly distributed deviation points

To rely only on the above requirementscan easily create errors. For example, for 6calibration points it is always possible to fita polynominal of the fifth degree such thatthe calculated curve runs through allcalibration points. However you usuallyreceive swinging curves, i.e. curves whichhave partially increasing slope, which isphysically meaningless. The followingrequirements should be fulfilled inaddition:

• The slope of the curve should bephysically meaningful. You can view thederivative of the calibration curve in therespective column of the editor sectionand as graphical information by selectionfrom the Compare list

• The slope of the calibration curveshould be highly negative for small andlarge elution volumes while there shouldbe a broad region with relativelyconstant value within.

11. Choose [file | print] to make a printout ofthe calibration table and graph.

12. Save the calibration file using the [Save As]dialog from the File menu and give adescriptive name for this calibration file,e.g. AK_Calib.cal.

41

VIII. Interactive calculation of molecular weight data

1. In ChemStation Data Analysis load themethod AK_SAM.M

2. Load a data file obtained for the broaddextran test sample.

3. Select [GPC], then [GPC Settings] andload the calibration file AK_Calib.cal youhave saved in the end of chaptercalibration

4. Press OK and then Save methodAK_SAM.M, the calibration fileAK_Calib.cal is now saved with the methodAK_SAM.M

5. Next select the [GPC] menu and then[Calculate GPC Results]

The file is transferred to the GPC dataanalysis software which displays the GPC-Addon top-level window

Figure 9:Top-level window of the Agilent GPC data analysissoftware for broad dextran test sample

a) Defining the baseline

In the raw data-(R) part define next thebaseline by moving the red triangles to thecorrect positions. Ideally the baselineshould be taken as a straight line betweenthe elution prior to the size exclusion limit,about 5 mL, and that after the last impuritypeak, i.e. areas in which no elution cantake place in an ideal GPC-SECseparation(1,3), refer to figure 9. For settingthe baseline correctly make sure byzooming that there is

• no peak elution at the position of thered triangles

• no cut off of the polymer peak


42

Figure 10:Raw data window for broad dextran test sample withbaseline markers set prior to size exclusion and lastimpurity peak

Fig. 11:Raw data window for broad dextran test sample withbaseline markers set at peak start and end

If this drawing of the baseline does notwork properly – e.g. peaks from previousanalysis elute prior to the size exclusionlimit, or the baseline is not sufficientlystable after the last impurity peak – youmay set the baseline markers exactly in theposition where the polymer peak startsand ends. For better seeing this you haveto zoom in.

Baseline setting as shown in figure 11 hasthe drawback of less flexibility for settingthe integration limits in the Elugram (E)window (refer to part b) but is often themore practical approach.

baselinemark

baselinemark

43

Fig. 12:Elugram window showing correct position of integrationlimits

b) Defining the integration limits

In the Elugram window the baselinecorrected raw data are presented.In the Elugram window the integrationlimits for calculation of the molecularweight distribution and the molecularweight averages have to be set with the redmarkers. Position them exactly before thestart and exactly at the end of the peak.You will have to zoom in for better seeingthis. This step is typically not needed if youset the baseline as shown in figure 11.

During the optimization of baseline andintegration limits study the drasticinfluence of the position of the triangles onthe molecular weight results. You canchange especially the Mn -value drasticallywith the set points. This strong influence ofthe integration limits and the baselinesetpoints on the molecular weight data isone of the reasons for the limited precisionand reproducibility of the molecularweight averages.

A typical value for the Mw-value of thebroad polystyrene sample under thechosen conditions is:

Molecular weight Typical GPC value Limits (± 20%)

Mw 32000 25600 - 38400

The Mw-data should be in a range of± 20% within the typical value. Try to comeclosest to the reference value. If youcannot succeed refer to Part CAppendices, chapter I. Troubleshooting.


44

Hardware Parameters Software Parameters

• Column performance • Accuracy of calculation procedures• Flow precision • Quality of baseline setting• Temperature precision • Quality of setting the calculation• Baseline stability start and end marks• Injection volume precision • Quality of calibration curve• Quality of the standards • Number of data points

• Possibility to use an internal standardcorrection for flow rate changes

The lab to lab reproducibility of molecularweight data depends on several hardwareand software parameters as the tablebelow shows:

Because of the many influencingparameters lab to lab reproducibility isoften poor. R. Bruessau reports (1) of anEuropean round robin test, whichresulted in differences from lab to lab forMn of ± 16 % and for Mw of ± 9%. Similardata were reported from a Japanese roundrobin test(2). In both round robbin studiesonly experienced GPC laboratoriesparticipated!

6. Highlight the [Mass Distribution(M)]window, then select the [Raw Data] menuand press [Print].You will get a single page report printedconsisting of the mass distribution, themain method parameters and in thebottom the molecular weight data.

7. The molar mass distribution should looklike the one shown in figure 13.

45


Fig. 13:Typical molecular weight distribution of the broaddextran sample

8. Determine the molecular data in the sameway for the second data file obtained withbroad dextran test sample (yellow vial).

46

IX. Automatic calculation of molecular weight data

1. Make sure that the system is still runningusing the above conditions and thatmethod AK_SAM.M is loaded. In the [GPCSettings] window the calibration fileAK_CALIB.CAL is activated.

2. Load the last data file obtained for thebroad dextran test sample.

3. The integration parameters in AK_SAM.Mneed to be optimized for the peak shapeobtained for the broad dextran with yoursystem. The GPC data analysis part of theChemStation calculates in a sequenceautomatically the molecular weight databetween the start mark of the 1st

integrated and the end mark of the lastintegrated peak. It is therefore extremelyimportant that

• Only the polymer peak is integrated(refer to figure 14)

• The start and end marks for the polymerpeak are placed exactly where the peakstarts and ends (zoom in several timesfor better seeing this, refer also tochapter VIII and figure 11)

• For optimizing the start and end marksuse (refer to the ChemStation help text):

Remarks:

• Slope sensitivity Big impact on start and end mark(use values as 0.01 or 1000 for studying th influence)

• Peak width Big impact on start and end mark(use values as 0.01 and 10 for studying the influence)

• Negative peaks On Negative peaks are integrated well andbaseline is not drawn to minumum of negativepeaks

• Integration On/Off For suppression of peaks of no interest

• Baseline Now Forces integrator to reset baseline

In this chapter is explained how to set upthe system for a completely automaticanalysis consisting of data acquisition, GPCdata analysis and reporting based on theabove developed method. An Agilent 1100Series GPC analysis system with degasser,pump, autosampler, thermostatted columncompartment and refractive indexdetector is required.

a) Preparation and start

47

Change the above parameters until thestart and end marks are optimized. Correctdrawing of the baseline itself within thestandard Data Analysis is not important.The GPC data analysis itself draws thebaseline arbitrarily from start to end mark.

Fig. 14:Example for broad test sample with correctly setintegration marks

4. Go into the [GPC Settings…] and

• Change the Report Settings from“Interactive Screen Review” to“Print Results”

• Configure a report consisting of, e. g.,“Graphical MWD”, a “Method”- and a“Results” part

5. Save the method again and set up asequence consisting of 6 injections of thebroad dextran test sample


b) Evaluation

1. You should have received automaticallythe following printouts for each run

• A standard HPLC report• A single page GPC report consisting of

one “Graphical MWD”-, one “Method”-and one “Results” part

2. In the ChemStation Data analyis make anoverlay of the six analysis

3. The overlay should look visually like asingle chromatogram, at least in thepolymer region.


48

You may have to discard the 1st analysis toobtain repeatable chromatograms.

3. Calculate the Standard Deviation (SD) andRelative Standard Deviation (%RSD) forthe obtained Mn- and Mw-values using thefollowing formulas or your pocketcalculator:

Fig. 15:Overlay of chromatograms obtained fromautomatic analyses

4. The relative standard deviations should be

• For Mn < 10 %• For Mw < 7 %

If this is not the case refer to Part C,chapter I. Troubleshooting. For moreinformation on precision of molecularweight data achievable with optimized,state of the art equipment refer toliterature 5)

49

X. Applications

Fig. 16:Quality control of dextran

a) Quality control of dextran

The chromatogram shows the analysis of adextran used in cosmetic formulations.

The initial analysis with just onePL aquagel-OH Mixed column indicatedalready with a shoulder in the mainproduct presence of impurities. Throughthe addition of one PL aquagel-OH 30column the resolution could be furtherincreased thus clearly proving theimpurity.

Columns: PL aquagel-OH Mixed, 7.5 x 300 mm, 8 µm(Agilent Part No. 79911GF-MXA)in series withPL aquagel-OH 30, 7.5 x 300 mm, 8 µm(Agilent Part No. 79911GF-083)

Mobile phase: WaterColumn temp.: 25 °CFlow rate: 1 mL/minDetector: Refractive Index DetectorCalibrant: Polyethylene oxide EasyCal standards

in vials for calibration


50

b) Quality control of polyethoxylate

Polyethoxylates from trimethylol-propaneare used in polymer industry for theproduction of protective coatings, resins,elastomers and as intermediates for thesythesis of other products.

The chromatogram shows the analysis of aproduct with a very high molecular weightpart eluting between 10 and 11 minutesand a lower molecular weight part elutingbetween 12.5 and 17.5 minutes.Because of the bimodal molecular weightdistribution the PL aquagel-OH Mixedcolumn was used in series with onePL aquagel-OH 30 column.

Fig. 17:Quality control of polyethoxylate

Columns: PL aquagel-OH Mixed, 7.5 x 300 mm, 8 µm(Agilent Part No. 79911GF-MXA) in series withPL aquagel-OH 30, 7.5 x 300 mm, 8 µm(Agilent Part No. 79911GF-083)

Mobile phase: WaterColumn temp.: 25 °CFlow rate: 1 mL/minDetector: Refractive Index DetectorCalibrant: Polyethylene oxide EasyCal standards

in vials for calibration

A B

Summary of results:

Part Mn Mw

A 1920000 2002000B 4000 6000

51

Column: 2 x PL aquagel-OH 30 in series, 7.5 x 300 mm, 8 µm(Agilent Part No. 79911GF-083)

Mobile phase: 0.2 M NaNO3, NaH2PO4 , pH 7Column temp.: 25 °CFlow rate: 1 mL/minDetector: Refractive Index DetectorCalibrants. Polysaccharides (Part No. 1535-4546)

(10 x 0.2 g)

c) Analysis of heparins

Heparins are mucopolysaccharides formedby the reaction of D-glucoseamines and D-glucuronicacid. They are widely used asanticoagulants to prevent thrombolysis incase of hyperlipidemia, arteriosclerosis,blood transfusions and operations. Theirproperties depend on the molecularweight and the molecular weightdistribution which can be fast andprecisely controlled by aqueous GPC-SEC.

Fig. 17:Analysis of heparins


Summary of results:

Part Mn Mw

A 22500 28200B 9350 10770

52

Columns: 3 x PL aquagel-OH 30 in series, 7.5 x 300 mm, 8 µm(Agilent Part No. 79911GF-083)

Mobile phase: 0.2 M NaNO3 , NaH2PO4 , pH 7Column temp.: 25 °CFlow rate: 1 mL/minDetector: Refractive Index Detector

d) Analysis of polyvinyl alcohol

Polyvinyl alcohols are industriallysynthesized by the catalytic reaction ofpolyvinyl acetates with alcohols, typicallymethanol. Due to properties as excellentbiological degradeability, water solubility ,toxilogical harmlessness they are widelyused as emulgators, binding agents, inadhesives, salves, haircream. Theproperties can be varied with themolecular weight distribution and themolecular weight which ranges from 20000to 100000 g/mol.

Fig. 18:Analysis of polyvinyl alcohol

For more application examples refer to“Polymer and Hydrocarbon ProcessingSolutions Guide”, Agilent publicationnumber 12-5968-7020E.

53


55

Part C


AnalysisAppendices

56

I. Troubleshooting

Problem:

1. Unstable retentiontimes

2. Consecutive injectionsfrom the same standardsand/or the broad testvial do not overlaywell

3. Molecular weight dataMw not within ± 20%of typical value

Possible Causes:

Column not well conditioned

Leak

Air bubbles in eluent

Defective pump seals

Column temperaturenot stable

Column adsorbs polymer,column not well conditioned

Pump flow not stable

Column temperaturenot stable

Injection volume not precise

Standards/samples notcompletely dissolved

Flow precision not good:the system must easilyperform the factoryspecification < 0.3%

Column temperatureprecision not good:thermostat should perform thespecification: ± 0.15 °C

Detector baseline: drift andnoise too high, exact settingof baseline and calculationstart/end marks not possible

Quality of standards

Quality of broad sample

Defective column

Solution:

Flush with eluent

Search for leak and remove it

Defective degasser pump asindicated by red error lamp

Replace seals

Check column thermostat

Flush with eluent and injectagain until o.k., GPC-SECcolumns often adsorbpolymers until an equilibrium isachieved

Refer to 1.)

Refer to 1.)

Rotor seal defectiveMetering seal

Refer to chapter IV.Preparation

Leak (refer to 1.)

Air bubbles in eluent(refer to 1.)


Switch thermostat off and on

Perform thermostat test fromDiagnosis part of ChemStation

Flush sample and referencecell for 10 minutes

Activate “Recycle” from the RIcontrol menue for severalminutes


Optical unit temperature notstable

Prepare standards newly foreach calibration

Prepare samples newly formeasurement

Handle column as described inthe PLgel column user’s guide

Determine plate number andsymmetry according to testchromatogram supplied withthe column

57

Part C Appendices

Possible Causes:

Baseline not set correctly

Calculation start and endmarks not set correctly

Calibration curve fit notoptimized

Wrong calibration curvein use

Flow precision not good:the system must easilyperform the specification< 0.3%. A typical, wellmaintained GPC system hasa retention time precision< 0.1%

Column temperatureprecision not good:thermostate should performthe specification: ± 0.15 °C

Detector baseline: drift andwander too high, repeatablesetting of baseline start/endmarks not possible

Baseline start and endmarks not set repeatable

Column impurities fromprevious application elute

Degradation of standard/test sample

Baseline wander ofrefractive index detectortoo high

Solution:

Repeat calculation asexplained in chapter VII

Optimize curve fit as explainedin chapter VI

Select correct calibrationcurve

Leak (refer to 1.)

Air bubbles in eluent(refer to 1.)


Switch thermostat off and on

Perform thermostate test indiagnosis and OQ/PV parts ofChemStation

Flush sample and referencecell for 10 minutes to get airbubbles out, use recycle modefor several minutes for betterflushing


Optical unit temperature notstable

Optimize integrationparameters as outlined inchapter IX. AutomaticCalculation of MolecularWeight Data.

Columns for GPC-SEC withorganic eluents:clean column by flushingthoroughly with THF

Columns for SEC with aqueouseluents:clean column by flushingthoroughly with water. If thisdoes not help, add up to 50%methanol

Prepare standards and testsample always newly

Purge reference cell

Activate recycle mode forbetter flushing

Problem:

3. Molecular weight dataMw not within ± 20%of typical value

4. Relative standarddeviationfor Mn not < 10%,for Mw not < 7%

5. Unknown peaks elutebefore the volume of totalexclusion and afterthe volume of totalpenetration

58

II. Literature

(1)R. J. Bruessau, “Experiences withInterlaboratory GPC Experiments”, Maromol.Symp. 110, 15-32 (1996)

(2)The Japan Society for Analytical Chemistry,Bunseki Kagaku 44(6), 497-504 (1995)

(3)DIN 55672-1 Gelpermeationchromatography(GPC), Teil 1: Tetrahydrofuran (THF) alsElutionsmittel, Berlin 1994

(4)ISO/TC35/SC N 950 = ISO CD 13885 1994-05-8Gel permeation chromatography

(5)H. Goetz, H. Schulenberg-Schell, “ImprovedPrecision of Molecular Weight Data usingadvanced GPC procedures”, submitted forpublication“Int. Journal of Polymer Analysis andCharacterization”.

Recommended books

(1)“Size Exclusion Chromatography”, edited byB. J. Hunt and S. R. Holding, Blackie andSon Ltd.

(2)“Handbook of Size Exclusion Chromatography”edited by Chi-san Wu, Marcel Dekker inc.

59

Part C Appendices

III. Acknowledgement

The Consumables and AccessoriesBusiness Unit of Agilent Technologiesrecognizes the efforts of Heinz Goetz,Pharmaceutical Business Unit, in thedevelopment of these kits.

For more information on our products visitour Agilent home page on the worldwide web at:http://www.agilent.com/chem/supplies

Printed in Germany 3/00Agilent Part No. G2182-90100

inorganic anions - agilent · 2016. 8. 29. · title: inorganic anions author: willi subject:...

Documents