Investigation of robustness for supercritical fluid chromatography separation of peptides: Isocratic vs gradient mode

We investigated and compared the robustness of supercritical fluid chromatography (SFC) separations of the peptide gramicidin, using either isocratic or gradient elution. This was done using design of experiments in a design space of co-solvent fraction, water mass fraction in co-solvent, pressure, and temperature.

jou rn al h om ep a g e : w w w e l s e v i e r c o m / l o c a t e / c h r o m a

Martin Enmarka,b, Emelie Glennea, Marek Le´skoa,c, Annika Langborg Weinmannd,

Tomas Leeke, Krzysztof Kaczmarskic, Magnus Klarqvistd, Jörgen Samuelssona,∗,

Torgny Fornstedta,∗

a Department of Engineering and Chemical Sciences, Karlstad University, SE-651 88 Karlstad, Sweden

b Pharmacognosy, Department of Medicinal Chemistry, Uppsala University, Biomedical Centre, Box 574, SE-75123 Uppsala, Sweden

c Department of Chemical and Process Engineering, Rzeszów University of Technology, PL-359 59 Rzeszów, Poland

d Early Product Development, Pharmaceutical Sciences, IMED Biotech Unit, AstraZeneca, Gothenburg, Sweden

e Medicinal Chemistry, Respiratory, Inflammation and Autoimmunity, IMED Biotech Unit, AstraZeneca, Gothenburg, Sweden

Article history:

Received 27 April 2018

Received in revised form 1 July 2018

Accepted 5 July 2018

Available online 10 July 2018

Keywords:

SFC

Peptide

Gramicidin

Robustness

Method transfer

Water

Weinvestigatedandcomparedtherobustnessofsupercriticalfluidchromatography(SFC)separations

ofthepeptide gramicidin,usingeitherisocraticorgradientelution.Thiswas doneusingdesignof experimentsinadesignspaceofco-solventfraction,watermassfractioninco-solvent,pressure,and temperature.Thedensityoftheeluent(CO2-MeOH-H2O)wasexperimentallydeterminedusinga Corio-lismassflowmetertocalculatethevolumetricflowraterequiredbythedesign.Forbothretentionmodels, themostimportantfactorwasthetotalco-solventfractionandwatermassfractioninco-solvent Com-paringtheelutionmodes,wefoundthatgradientelutionwasmorethanthreetimesmorerobustthan isocraticelution.Wealsoobservedarelationshipbetweenthesensitivitytochangesandthegradient steepnessandusedthistodrawgeneralconclusionsbeyondthestudiedexperimentalsystem

Totesttherobustnessinapracticalcontext,boththeisocraticandgradientseparationsweretransferred

toanotherlaboratory.Thegradientelutionwashighlyreproduciblebetweenlaboratories,whereasthe isocraticsystemwasnot.Usingmeasurementsoftheactualoperationalconditions(notthesetsystem conditions),theisocraticdeviationwasquantitativelyexplainedusingtheretentionmodel.Thefindings indicatethebenefitsofusinggradientelutioninSFCaswellastheimportanceofmeasuringtheactual operationalconditionstobeabletoexplainobserveddifferencesbetweenlaboratorieswhenconducting methodtransfer

©2018TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense

(http://creativecommons.org/licenses/by/4.0/)

1 Introduction

The separation of therapeutic peptides has long been an

important application area for chromatography, particularly

reversed-phaseliquidchromatography(RPLC)[1 Withgrowing

interest in supercritical fluid chromatography (SFC) for

analyz-ingandpurifyingsmallmolecules(i.e.molecularweights<1kD)

[2,3 severalauthorsfrombothacademiaandindustryhavealso

startedtoinvestigatehowSFCcouldbeusedtoanalyzeandpurify

peptides[4–13] Whilethequality-by-design(QbD)paradigmis

firmlyestablishedinliquidchromatography[14],itisnotsimilarly

∗ Corresponding authors.

E-mail addresses: Jorgen.Samuelsson@kau.se (J Samuelsson),

Torgny.Fornstedt@kau.se (T Fornstedt).

establishedinSFC,probablybecauseSFCislessrobustthanliquid chromatography[3 Somestudieshaveinvestigatedthe robust-nessofSFCseparationmethodsinthecontextofmethodtransfer andbyinvestigatingtherobustnessinadesignspace[15] Thesmallbut growingbody ofstudiestreating theSFC sep-aration of peptides [4–13] has investigated a limited number

of peptides, for example, gramicidin D [6,12,13], leucine-enkephalin[4–6,10],methionine-enkephalin[4–6,10],angiotensin

I [4 angiotensin II [4–6], cyclosporin analogs [7 beta-methylphenylalanine [11], oxytocin [10], bradykinin [4,10], Pro-Leu-Glyamide[4 sauvagine[4 leupeptide[4 urotensinII [4 sulfomycin[8 cyclicpeptides[16],andcustomacidicandbasic linearuncappedpeptides[9

Most studies have used traditional liquid chromatographic stationary phases suchas silica [7,7,9 diol [8,9 C18 [4,9 2-ethylpyridine[4,5,9 cyano [6 and various chiral phases [11], https://doi.org/10.1016/j.chroma.2018.07.029

0021-9673/© 2018 The Authors Published by Elsevier B.V This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ).

Table 1

Properties of the gramicidin isoforms partially separated in the study.

formyl-X-Gly-Ala-Leu-

Ala-Val-Val-Trp-Leu-Y-

Leu-Trp-Leu-Trp-ethanolamine

* Specified by vendor.

** Not specified by vendor.

as well as polymer-based phases such as divinylbenzene [10]

or poly(styrene–divinylbenzene) [12,13] Eluents are typically

CO2 modified with acetonitrile/water [5,8 acetonitrile [4,11],

methanol[4,6–9,11],ethanol[7,11],andisopropylalcohol[7,11]

towhichacidicorbasicadditivessuchastrifluoroaceticacid(TFA)

[4,5,9 2,2,2-trifluoroethanol(TFE)[6 ammoniumacetate[4,6

aceticacid[6 andisopropylamine[6,8]areadded.Severalstudies

haveinvestigatedthemodificationofco-solventswithwater,and

founditsadditionnecessarytoachieveresolutionortoimprove

peakshape[4,6,9 Moststudieshaveusedgradientelution,but

somehavealsoinvestigatedtheisocraticelutionmode[11]

Duetothesmallchemicalspaceinvestigated,itisdifficultto

drawgeneralconclusionsastothefeasibilityofusingSFCfor

pep-tideanalysisandpurification.However,severalofthementioned

studiesdid investigatethe effects of thestationary phase,

elu-ent,andotheroperationalconditions,suchasbackpressureand

temperature[7 Clearly,amechanisticunderstandingofpeptide

separationinSFCislackingcomparedwithourunderstandingof

themuchmorematureRPLCtechnique[17–19]

Robustness is “a measure of [an analytical procedure’s]

capacitytoremainunaffectedbysmall,butdeliberatevariationsin

methodparametersandprovidesanindicationofitsreliability

dur-ingnormalusage”[20].Itiswellknownthatdeliberatevariations

inoperatingconditionsinSFCcangreatlyaffectseparation[21,22]

whichisanadvantageofSFCascomparedtoLCforimproving

selec-tivity;however,thisalsoaffecttherobustness.However,itisless

knownandunderstoodthatunintentionalvariationscanalsohave

amajorimpact,forexample,whenoperatingSFCinhighly

com-pressibleregionsorwhengeneralretentionmechanismsarepoorly

understood.Studyingtherobustnessofseparationsconductedin

thehighco-solventregimeofSFC,technicallyinsubcritical

condi-tions[23]cangivevaluableinsightintoareastypicallynotstudied

inSFCwhere theeluent ismoreLClikebecausethefluid

com-pressibilitydecreaseswithincreasingco-solventintheeluent.Most

studiesindicatethatworkingwithalargefractionofco-solventis

necessarytoelutepeptides

Beyazetal [24] systematically studiedtheeffects of

differ-ent instrumental and operating conditions on the precision of

retentiontimesforalargesetofsoluteselutedonC18using

ace-tonitrile/buffer/water.Theyconcluded,forexample,thatisocratic

elutionwasmoresensitivethanwasgradientelutionwhen

study-ingtheeffectsofvariationinthemobilephasecomposition.No

similarinvestigationhasbeendoneforSFC

Theaimofthisstudyistoinvestigateandcomparethe

robust-nessofpeptideseparationsconductedunderisocraticandgradient

conditions in SFC As a model compound, we studied the

lin-earunchargedpentapeptidegramicidinseparatedonapH-stable

hybridsilicacolumnusinganeluentcontainingCO2,water,and

methanol.Therobustnesswasinvestigatedbyevaluatingthe

vari-ationin theretentionfactor usingdesign ofexperiments (DoE)

by perturbing the most important operational conditions, i.e

varyingthetotal orinitialgradientfractionofco-solvent,water massfractioninco-solvent,pressure,andtemperature.Secondly, simulationsbasedontheexperimentaldata,butwithdifferent sen-sitivitiestotheperturbations,wereperformedinordertoreveal howtherobustnesswouldvaryforhypotheticalsolutesin gradi-entseparationsofdifferentgradientslopes.Finally,thepractical consequencesoftheobserveddifferencesinrobustnessbetween gradientandisocraticseparationswerequantifiedbytransferring theisocraticandgradientmethodstoadifferentlaboratory

2 Material and methods

2.1 Chemicals The mobile phase consisted of CO2 (99.99%) from AGA Gas

AB(Lidingö,Sweden),HPLC-grademethanolfromVWR(Radnor,

PA, USA), and water with conductivity of 18.2Mcm from a Milli-QPlus185waterpurificationsystemfromMerckMillipore (Darmstadt,Germany).Gramicidin(CAS#1405-97-6)fromBacillus aneurinolyticuswasobtainedfromSigma-Aldrich(St.Louis,MO, USA).Thislinearpeptidehasthesequenceformyl-

X-Gly-Ala-Leu-Ala-Val-Val-Trp-Leu-Y-Leu-Trp-Leu-Trp-ethanolamine, where X

canbeeitherValorIleandYeitherTrp(GramicidinA),Phe (Gram-icidinB),orTyr(GramicidinC)[25],andisthereforereferredto

ascomprisingtheX–Yisoformsofgramicidin(Table1).All Gram-icidinsamplesweredissolvedinneatMeOHtoaconcentrationof

1mgmL–1 2.2 Instrumentation

Inthisstudy,twodifferentanalyticalSFCsystems,ofthesame modelandmanufacturerwereused,butattwodifferentlocations: KarlstadUniversity(denotedLaboratory1)andatAstraZenecain Gothenburg(Laboratory2).TheLaboratory1systemwasaWaters UPC2(WatersCorporation,Milford,MA,USA)equippedwithaPDA detector.TheLaboratory 2systemwasalso aWaters UPC2 but connectedviaapassivesplitter(UPC2MSsplitter)toaPDA detec-toranda Waterssingle-quadrupolemassspectrometer(Waters SQD2)usingelectrosprayionizationinpositivemode.Bothselected ionmonitoringandscan mode(800–1000m/z;833Das−1)were used,respectively.InLaboratory1theUPC2instrumenthada sin-glestackconfigurationfrombottomtotopofpump,autosampler, convergencemanager(backpressureregulator),columnmanager andPDAdetector.InLaboratory2,theUPC2inLaboratory2had

atwo-stackconfigurationwithpump,autosampler,convergence managerinthefirststackandmake-uppump,columnmanager andPDAdetectorinthesecondstack.Theinnerdiameterofthe stainlesssteeland PEEKtubingfrominjectortocolumntoPDA

toconvergencemanagerwas0.18mmatbothlaboratoriesexcept fromthesplittertoconvergencemanageratLaboratory2werethe innerdiameterwas0.25mm.ThePDAflowcellvolumewas8.4␮L

atbothlaboratories

Themobilephaseflowtothemassspectrometerwassplitwith

apassivesplitteranddilutedwitha0.2mLmin–1mixtureof95/5% (v/v)methanol/10mMammoniumformate.Theextracolumn vol-umewas measuredfrom the retention time of an injectionof

1mgmL–1gramicidinwithazero-dead-volumeunioninplaceof thecolumninbothsystems.Thedifferencewascompensatedfor

incomparisonsbetweenthetwosystems.Pressurewasmeasured usingtwomodelEJX530Aabsolutepressuretransmitters (Yoko-gawaElectricCorporation,Tokyo,Japan)connectedtothecolumn inletandoutletusingatee.AdataloggerfromPicoTechnology(St Neots,UK)wasusedtorecordthepressure

Thetotal andco-solventmassflows weremeasureddirectly afterthemobilephasemixerandbetweentheco-solventpump

low-flowCoriolismassflowmeter(BronkhorstHigh-TechB.V.,Ruurlo,

Netherlands),hereafterdenoted“CFM.”Thecolumnsusedwerea

2.5-␮mKromasil SFC-2.5-XT(100×3.0mm)(AkzoNobel,Bohus,

Sweden) and a 1.7-␮m Waters 2-picolylamine (100×3.0mm)

(WatersCorporation)

2.3 Procedure

2.3.1 Designofexperiments

A three-level, four-factor, central composite face-centered

experimentaldesignwiththreecenterpointswasusedto

inves-tigatehow the logarithmic valueof the retention factor ofthe

Val-Aisoformofgramicidinvarieswithtotalco-solventfraction

(MeOH+water),watermassfractioninco-solvent,pressure,and

temperaturefortheisocraticandgradientelutions,respectively

OnlytheKromasil SFC-2.5-XTcolumnwasinvestigated.Thelog

transform of the retention factor wasused because the

reten-tion generally hasa logarithmic relationship with the fraction

ofco-solventusedintheseparation[26].Thecentralcomposite

face-centeredexperimentaldesignmodelwasselectedinorderto

achievegood predictivepowerinthedesign space[27]as well

astoinvestigatepotentialquadratictermsandinteractionterms

betweenfactors.Intheisocraticelutionexperiments,thetotal

co-solventfractionwasidenticaltotheisocraticcomposition,andin

thegradientelutionexperiments,thetotalco-solventfraction

indi-catedtheconditionatthestart(andend)ofthegradient.Theset

designwasasfollows:thetotalco-solventfractionduringthe

iso-craticexperiments was30, 33,and 35 v/v%and theco-solvent

gradientwas28.3–61.3,30.0–65.0, and31.7–68.7 v/v%in5min

whentheretentiontimeswerefoundtobereasonable.Aftereach

changeofeluentcomposition,thesystemwasequilibratedforat

leastonehour.Thewatermassfractioninco-solventwas1.2,5,

and8.7w/w%.Theset,backpressurewas110,130,and150bar

Thetemperaturewas30,45,and60◦C.Duetothenatureofmixing

abinaryco-solvent[28,29]withCO2,thesetvolumetricfraction

ofco-solventwasnotusedbutratherthemeasuredmassfraction

[30].The design wasrescaled for theactualandmeasured

val-uesoftheco-solventfraction,watercontent,andpressure.2␮L

injectionsof1mgmL–1 gramicidininneatMeOHweremadeat

leastinduplicateforeachexperimentalcondition.Chromatograms

wererecordedat220nm.Retentiontimeswereestimatedfrom

peakapexandnormalizedtoretentionvolumesusingthe

mea-suredmassflowanddensity(seesection2.3.2).Thevoidtimewas

obtainedfromtheinitialbaselinedisturbanceandwasalso

normal-izedtovoidvolume.Theaverageofallvoidvolumeswasusedin

calculatingeachretentionfactor.Multiplelinearregressionsofthe

log10-transformedretentionfactorswereperformedusingMODDE

11(Umetrics,Umeå,Sweden)witha95%confidenceleveland

non-significantfactorsweremanuallyremoved

2.3.2 Characterizingtheexperimentalconditions

Asfactorsfortheexperimentaldesign,thetotalco-solventmass

fraction,watermassfractioninco-solvent,columntemperature,

andaveragecolumnpressurewereused.Theco-solventfractions

weremeasuredusingtheCFM.Thearithmeticmeanofthecolumn

inletandoutletpressuresforeachisocraticandgradientcondition

wasusedasthepressurefactor.Theinstrumentsettemperature

wasusedasinputtotheexperimentaldesign,asseveralof our

studieshaveindicatedthatourinstrumentsettemperatureisvery

accurate[22,30].ThemassfractionofwaterinMeOH,takenfrom

thegravimetricpreparationofco-solvents,wasusedasinputtothe

experimentaldesign

Tocalculatethevolumetricflowrate,thedensityoftheeluentis

required.However,toourknowledgeitisimpossibletoaccurately

calculatethedensityoftheternaryCO -MeOH-H Ofluidusedhere

Therefore,directdensitymeasurementusingtheCFMwas eval-uatedandperformed.Themainchallengewasthatthepressure andtemperatureinsidetheCoriolisflowcellmustbeidenticalto thoseinsidethecolumn.Thiswasachievedbyremovingthe col-umnandsettingtheback-pressureregulatorsothatthecolumn averagepressureswereachievedintheCFM.Thetemperaturewas adjustedbysimultaneouslyincreasingtheflowrateandtheset col-umnoventemperatureuntilthedesiredtemperatureintheCFM wasobtainedandstabilized.TubingfromtheUPC2totheCFMwas insulatedtominimizeheatloss

To plotcontourplots and calculate densities otherthan the experimentalmeasureddatapoints,seeSupplementaryDataTable S1; theexperimentaldatawerefitted toa second-order multi-polynomialequationwithinteractiontermsusingMODDE11 Theaccuracyofthesedensitymeasurementswasfirst evalu-atedbycomparingtheoreticalandmeasureddensitiesusingpure

CO2atthreesetbackpressures(110,130,and150bar)andthree temperatures(30,45,and60◦C)at3mLmin−1.Thetheoretical den-sitywascalculatedusingNISTReferenceFluidThermodynamicand TransportpropertiesDatabaseversion9.1(REFROP)[31]withthe measuredarithmeticmeanpressureandmeasuredtemperatureas inputs,seeSupplementaryDataTableS2

Allpressureanddensitymeasurementswereconducted sepa-ratelytominimizeextracolumnvolumes

2.3.3 Methodtransferexperiments Thesame2.5-␮mKromasilSFC-2.5-XTusedfortheDoEin Lab-oratory1wasinstalledinLaboratory2andthesamesetmethod conditionswereusedaswhenrunningtheexperimentaldesigns center-pointexperimentsintheisocraticandgradientelutions.The totalmassflow,co-solventmassflowandaveragecolumnpressure weredeterminedatbothsites

3 Results and discussion

TheretentionbehaviorinSFCofthemainisoformofgramicidin, Val-A,wasinvestigatedintheisocraticandgradientelutionmodes usingamixtureofMeOHandwaterasco-solventsatdifferent tem-peraturesandpressures.Thegoalwastouseaquantitativemodel

oftheretentionfactortocomparetherobustnessoftheseparation systemineitherelutionmodewithinthedefineddesign space Experimentaldatawerealsoextrapolatedtogivegeneralinsight intotherobustnessoftheisocraticand gradientelution separa-tionsystems.Tocalculatetheretentionvolumeintheexperimental space,theeluentdensitywasdeterminedusingtheCFM.Finally, theseparationsystemwastransferredtoadifferentlaboratoryto evaluatethepracticalimplicationsoftransferringamoreorless robustseparationsystem

3.1 Retentioncharacteristicsofgramicidin Thegoalofthescreeningwastofindasatisfactoryseparation systemandtofindsuitableboundariesfortheexperimentaldesign Initialscreeningofthechromatographicbehaviorofgramicidin anditsisoformswasdoneonhybridsilicaand2-picolylamine sta-tionaryphasesusingMeOH/waterastheco-solvent.Fig.1presents thechromatogramfroma2-␮Linjectionof1.0mgmL–1gramicidin separatedonthehybridsilica(Fig.1a)and2-picolylamine(Fig.1b) columns.Frommassspectrometricdata,theretentionorderonthe hybridsilicaphasewasfoundtobeIle-B,Val-B,Ile-C/Ile-A,and Val-C/Val-A(Fig.1c)andonthe2-picolylaminephasetobeIle-B/Val-B, Ile-C/Val-C,Ile-A,andVal-A(Fig.1d).The2-picolylamine station-aryphasemanagedtoresolveeacharomaticisoformbutnotthe aliphaticforms,exceptforIle-AandVal-A.Thehybridsilica sta-tionaryphase,ontheotherhand,managedtoresolvethealiphatic isoformsbutwaslessabletodifferentiatebetweenthearomatic

Fig 1. Analytical injections of gramicidin (a, c) on the hybrid silica column (Kromasil SFC-2.5-XT) using 5.00-min gradient elution of 28–62 v/v% at 110 bar and 60 ◦ C and (b, d) on the 2-picolylamine column using 5.00-min gradient elution of 23–57 v/v% at 110 bar and 60◦C Top row shows 220-nm UV traces of 2-␮L injections of 1 mg mL –1

gramicidin Bottom row shows selective ion traces of all gramicidin isoforms for [M + 2 H] 2+ fragments.

forms.Theseresultsareasexpectedconsideringthenatureofthe

hybridsilicaandthe2-picolylamineligand.Furtherstability

exper-imentsusingthe2-picolylaminecolumnrevealedanon-reversible

retentiondriftwhenvaryingtheamountofwater,sothiscolumn

wasnotusedinfurtherstudies(datanotshown)

Addingwatertothemethanolco-solvent[32,33], wasfound

tosignificantlyaffecttheretentionandpeakshapeinthecaseof

gramicidin(Fig.2).Toinvestigatewhetheraddingwater tothe

eluentresultedinacontinuousordiscontinuouschangein

reten-tionand/orpeakshape,thefirstinjectionswereperformedwith

neatmethanolonnewcolumnsusingtheisocratic(Fig.2a,b)and

gradient(Fig.2c,d)elutionmodes.Followingtheneatmethanol

experiments,injectionsweredoneat1.2,5,and8.7w/w%water

addedtotheco-solvent.WhilethesolubilityofwaterinneatCO2in

supercriticalconditionsisgenerallybelowamolarfractionof0.01

[34],itissignificantlyhigherwhenthewaterisaddedtogetherwith

methanol[35].Byincreasingthewatercontentoftheeluent,the

apparenttailingofthemainpeakdecreasesinsemi-analytical

con-ditions(Fig.2a,c)andisconsiderablyreducedinsemi-overloaded

conditionsinboththeisocraticandgradientelutions(Fig.2b,d)

Toconclude,wefoundtheretentiononthehybridsilica

col-umntobereproducible andabletoseparatealiphaticforms of

gramicidin.Wealsofoundthatwaterreducedtheretentionfactor

and considerably reduced the peak tailing, especially in

semi-overloadedconditions.Addingwater totheeluentinthis range

didnotinduceanydiscontinuousorunexpectedbehaviorsinthe

retentionorpeakshape

3.2 Measurementofdensitytoestimatevolumetricflow

ToevaluatetheestimationofdensityusingtheCFM,the

den-sityofneat CO2 wasmeasuredovertherange of 30–60◦C and

134–175bar,inwhichtheCO2densityvariesfrom530to871kg

m−3(SupplementaryDataTableS2).Comparingthemeasuredand

calculated(REFPROP)densitiesshowedthattherelativedifference neverexceeded0.4%.ThisindicatesthatCFM shouldbeableto accuratelymeasuretheeluentdensity

BecauselittleisknownofthepropertiesoftheCO2-MeOH-H2O eluentsusedhere,thedensitywasmeasuredatallexperimental conditions(SupplementaryDataTableS1).Thesedatawerethen fitted toa second-order multi-polynomial equationwith inter-actionterms tointerpolatedensitiesin otherconditions.It was possibletofindanacceptablecorrelation(R2=0.79atthe95% con-fidencelevel)betweenthefactorsandthemeasureddensity Fig.3a–cplotsthedensityvariationasafunctionof tempera-tureandpressureforaco-solventfractionof31.5w/w%with1.3 (a),5 (b),and 8.7 (c) w/w% water intheco-solvent Ascanbe seen,thedensityvariesonlyslightlywithpressureand temper-ature,andaddingwatertotheeluentonlyslightlyincreasesthe densityofthemobilephase.Thismeansthat,fromadensity per-spective,thesystemisratherinsensitivetochangesintemperature, pressure,andthefractionofwateraddedtotheeluent.Littleis knownofthesysteminvestigatedhere,sowe cancomparethe resultsusingacalculatedCO2-MeOHmixturewithahighMeOH fractionintheeluent.ThedensityofaCO2-MeOHfluidatthe cen-terpoint(68.5/31.5w/w%co-solventfraction,5w/w%H2O,45◦C, and163.3bar)wasmeasuredtobe844±8kgm−3,whileitwas calculatedtobe843kgm−3usingREFPROP,indicatingthatwater haslittleeffectonthedensityoftheeluent

Fromthecorrelationof pressuretodensityatconstant tem-peratureand constant fractionsof co-solventandwater, it was alsopossibletodeterminehowdensityvariedinsidethecolumn duringaseparation.Fig.3 plotsthedensityvariationalongthe column assuminga linear pressuredrop at thecenter point in theexperimentaldesign.Thedensityalongthecolumnvariedby approximately1.5%fromcolumninlettocolumnoutlet(Fig.3d), meaningthatitisreasonabletousetheaveragedensityto deter-minetheaveragevolumetricflowrate

Fig 2.Analytical (a, c) and semi-preparative (b, d) injections of gramicidin at 0, 1.2, 5.0, and 8.7 w/w% water in MeOH co-solvent: (a, b) isocratic elution conditions at the center point of the DOE experiments, 35 v/v% co-solvent, 130 bar BPR, and 45 ◦ C; (c, d) gradient elution conditions at the center point of the DOE experiments, 30–65 v/v% co-solvent in 5 min, 130 bar BPR, 45 ◦ C Injections were 2 ␮L, 1 mg mL –1 (a, c); and 2 ␮L 20 mg mL –1 (b, d).

Fig 3.Density variation in the experimental design space: isopycnic plots for 1.3, 5, and 8.7 w/w% water in MeOH at the isocratic center point of 31.5 w/w% over the studied pressure and temperature range (a–c) Plot (d) shows the density profile along the column as a function of a linear pressure drop in the isocratic center point.

Table 2

Coefficients (scaled and centered) of design of experiments describing the retention factor of the Val-A isoform of gramicidin; numbers scaled by a factor of 10 2 , 95% confidence level C tot/init is the co-solvent fraction during isocratic elution or the initial fraction in gradient elution.

Log(k Val-A ) gradient b −6.69 ± 0.81 −3.75 ± 0.66 −2.00 ± 0.94 2.28 ± 0.66 1.45 ± 1.18 1.76 ± 0.692 −1.41 ± 1.00 75.9 ± 0.975

a Q 2 = 0.920, R 2 = 0.957.

b Q 2 = 0.942, R 2 = 0.973.

* Not significant.

Thedensitydropoverthecolumncouldbefurtherreducedby

operatingthesystematmuchlowerflowrates,asrecommended

earlier[36],butbothcolumnefficiencyandseparationtimewill

sufferforthisslightimprovement

Inthisinvestigation,theaveragevolumetricflowratevaried

between1.01 and 1.16mL min−1 over theentire experimental

design(SupplementaryDataTableS1),clearlyindicatingthe

impor-tanceof normalizing retention factorsbefore usingthem in an

experimentaldesign, as not doing sowould underestimatethe

retentiontimesbyupto16%andskewtheretentionmodel.Thishas

previouslybeensuggestedbyusandseveralotherauthors[29,37]

3.3 Robustnessofseparationsystem

Wechoseaface-centeredcentralcompositedesignforthe

pur-poseofquantitativelydescribingthevariationinretentionvolume

inordertoestimatetherobustnessoftheseparationsystem.All

lin-earfactorswerefoundtobesignificantaswellassomequadratic

andinteractionterms.TheircoefficientsarepresentedinTable2

Themodel determinedin theDoEdescribeshowtheelution

volumevarieswithchangesintotalco-solvent(w/w%),waterin

co-solvent(w/w%),pressure,andtemperaturewithinthe

exper-imentaldesign space.Usingthemodel,itispossibletoquantify

thesensitivityoftheseparationsystem,inthiscasetheretention

volumeofVal-A,toperturbationsintheco-solventfraction,water

massfractioninco-solvent,pressure,andtemperatureineitherthe

isocraticorgradientelutionmode.Themodelcoefficients(Table2)

indicatethattheisocraticseparationsystemis2.5,3.2,2.5,and1.4

timesmoresensitivetochangesinthetotalco-solventfraction,

watermassfractioninco-solvent,pressure,andtemperaturethan

inthegradientelutionsystem

Onewayofvisuallyrepresentingtherobustnessoftheisocratic

andgradientelutionsystemsispresentedinFig.4aforisocratic

conditionsandFig.4 forgradientconditions.Theplotrepresents

acontoursurfaceindicatingtherelativeerror,ER,oftheretention

factor,k,atanypositioninthedesignrelativetotheretentionfactor

atthecenterpoint,kref

ER=100· k−kref

kref



Theplotwasgeneratedtoinvestigatehowperturbationsinthe

twomostimportantfactors,totalco-solventfractionand water

massfractioninco-solvent,affecttheretention.Naturally,the

com-pleterobustnessoftheseparationsystemisrelatedtochangesin

anyfactorsinsideoroutsidetheexperimentaldesign.Startingat

theisocraticcenterpoint(Fig.4a,circle/cross),itisapparentthatif

thetotalco-solventfractioniskeptconstant,aperturbationofupto

approximately±5%inthewatermassfraction(observetherelative

changes,inthiscase4.75–5.25w/w%MeOH/H2O)intheco-solvent

wouldbeallowedifthemethodspecifiesthattheretentionfactor

canvaryby≤2%.Ifthetoleranceisincreasedto≤10%,a

pertur-bationofuptoapproximately–25/+30%inthewatermassfraction

wouldbepossible.Similarobservationscanbemadefor

pertur-bationsof total co-solventfractionwithaconstant water mass

fraction.Thesystemisleastrobustwhenbothtotalco-solvent

frac-tionandwatermassfractionaresimultaneouslyperturbedinthe

samedirection,becausebothfactorsaffecttheretentionvolumein thesamedirection.Fromthediagonalshapeofthecontoursurface,

itisalsoapparentthatifthefactorsaresimultaneouslyperturbed

inoppositedirections,itispossibletoperturbthesystem unknow-ingly,i.e.,maintaininganearconstantretentionfactorwhilehaving changedtheoperationalconditions.However,alargeperturbation

inthetotalco-solventfractionandwatermassfractionwouldalso alterthesystempressureandfurtherchangetheretentionfactor, makingtheinterpretationslightlymorecomplicated

Focusingonthegradientcenterpoint(Fig.4b,circle/cross),it

isapparentthatifthetotalco-solventfractioniskeptconstant,a perturbationofuptoapproximately±17%inthewatermass frac-tionwouldbeallowedifthemethodspecifiesthattheretention factorcanvary by≤2%.Ifthetoleranceis increasedto≤10%,a perturbationofuptoapproximately–80/+90%inthewatermass fractionwouldbepossible.Thecontoursurfacehasthesame char-acteristicsasinisocraticelution,meaningthataperturbationof bothfactorsinthesamedirectionoropposite directionswould maximizeorminimizetheresponseofthesystem,respectively Themostimportantconclusionisthatthegradientsystemisless sensitivetoco-solventorwaterperturbationsthanistheisocratic system

There are several possible origins of perturbations in the co-solventandwaterlevels,themostlikelytooccurand, simul-taneously,themosteasilymitigated isinaccuracy intheeluent preparation.BecauseMeOHisveryhygroscopic,anothersourceof perturbationistheaccumulationofwaterovertimedueto absorp-tionfromtheair.Changesintotalco-solventfractionaremuch moredifficulttoidentify,astheycouldresultfromdifferentpump performanceorpumpleakage,whichalsocouldbeaffectedby dif-ferentsystempressures.Thismatterisdiscussedfurtherinsection 3.5

3.4 Simulatedrobustnessofmodifiedseparationsystem The robustness of the studied separation system is a func-tionofthesensitivityoftheVal-Aisoformofgramicidintototal co-solventfraction, watermassfraction,pressure,and tempera-tureonthesilica-basedstationaryphase.Thisisdescribed,after removingallnon-significantterms,bythefollowingsecond-degree polynomial:

log10(k)=˛1P+˛2Ctot+˛3T+˛4CH 2 0+˛5T2+˛6PT

wherethecoefficients˛1to˛7andconstantˇarelistedinTable2

CH2O(w/w)isthewatermassfractionintheco-solvent,Ctot(w/w)

isthetotalfractionofco-solventintheeluent,T(◦C)isthe tempera-ture,andP(bar)isthepressure.Ifthepressureandtemperatureare keptconstantandwejustconsiderthewaterandtotalco-solvent, themodelcanbereducedto:

log10(k)=˛2Ctot+ (˛4+˛7T) CH 2 0+ (3) whereisaconstant.Usingthissimplifiedmodel,twoadditional separationsystemswereinvestigated:first,inwhichthe sensitiv-itytototalco-solventandwaterwasreducedtohalfthatofthe

Fig 4.Robustness plots of the separation system described by variation in the retention factor of the Val-A isoform, showing the two most important factors describing the system, i.e., total co-solvent fraction and water mass fraction in co-solvent Plot (a) shows the variation in isocratic elution mode and (b) in gradient elution mode The crossed dots indicate the center reference points in the isocratic and gradient elution modes where the relative error is zero.

Fig 5. Simulated robustness plots of isocratic elution based on the experimental system Plot (b) is a robustness plot using the simplified regression model (Eq 3); plots (a) and (c) represent theoretical systems less and more sensitive to the co-solvent and water fraction, by a factor of 2.

initialmodeland,second,inwhichthesensitivitywastwicethat

oftheoriginalmodel.Thepressureandtemperatureweresetto

bethesameasatthecenterpoint,andthesecontributionstothe

retentionwerehandledbyadjusting.Therobustnessplotsfor

isocraticconditionsarepresentedinFig.5a–c.Itisobviousfrom

theplotsthatthesystembecomesmorerobustasthesensitivity decreases(goingfromFig.5ctoa).Thisisunsurprisingandleadsto theconclusionthat,forrobustseparations,oneshouldavoid iso-craticseparationsifthesystemisverysensitivetochangesineluent composition

experimen-talsystemingradientelution,weevaluatedtheuseoflinearsolvent

strengththeory,[26,38]:

log10(k)=log10(k0)−S·Ctot (4)

wherek0istheretentionfactorusingneatCO2aseluent,Sisthe

sensitivitycoefficient,andCtot isthetotal co-solventfractionin

eluent.Thelinearsolventstrengththeory(LSS)wasdevelopedfor

liquidchromatography,buthaverecentlybeenreportedtodescribe

theretentionofsolutesinSFCforseparationsusinghighfractions

ofMeOHintheeluent[29,38]

Assumingthat Sis alinear functionofwater fractioninthe

eluent:

log10(k)=log10(k0)−(S0+S1·CH2O)·Ctot (5)

InclassicalSFCgradientexperiments,theco-solvent(mixture

ofwaterandMeOH)ismixedwiththeCO2.Thiswillresultinthat

boththeMeOHandwatercontentwillvarywithtimeinthe

elu-ent.Thisjustifiestoincludethewaterterminthegradientequation

ObservethatCH2Oisthewatermassfractionintheco-solventand

notthewaterfractionintheeluent.However,multiplyingCtotwith

CH2Owillgivethewaterfractionintheeluent.ParametersS0and

S1andlog10(k0)wereestimatedforeachisocraticsystem(i.e.,less

sensitive,normal,andmoresensitive)andaresummarizedin

Sup-plementaryDataTableS3.Assumingalineargradient,theretention

timecanbecalculatedasfollows:

tR= t0

G(Gkstart+1)+t0,

G=St0

tg

(6)

wherekstartistheretentionfactorforthestartingeluent

composi-tion,t0isthehold-uptime,tgisthegradienttime,Gisthegradient

steepnessfactor,andisthechangeintotalco-solventduring

thegradient

Ingradientelution,boththesensitivitytochangesinthe

co-solventaswellasthegradientslopeareimportant.Thesensitivity

washandledinthesamewayasintheisocraticcase:50%,100%,

and200%oftheinitialmodelsensitivity.Thegradientslopeswere

1%min−1, 7%min−1 (center point),and 13% min−1 changes in

co-solventfractioninthegradientrun.Thefirstshallowgradient

representsahigh-resolutionseparationandthelaststeepgradient

representsafasthigh-throughputscreeninggradient.Theresults

arepresentedinFig.6:(a)inthetoprow,1%min−1,(b)middle

row,7%min−1,and(c)bottomrow,13%min−1 gradientslopes

Fromthefigure,itisapparentthattherobustnessofthe

separa-tionincreaseswithgradientsteepness,goingfromtoptobottom

Theleftcolumn(I)ofplotsinFig.6representsaseparationthatis

lesssensitive,middlecolumn(II)normallysensitive,andright

col-umn(III)moresensitivetochangesintotalco-solventfraction.The

robustnessincreasesasthesensitivitytochangesintheco-solvent

decreases.Finally,wecanalsoobservethediagonalpattern(aI→bII

→cIII)illustratingthatthemoresensitivethesolute,thesteeperthe

gradientslopeneedstobeinordertomaintainsimilarrobustness

Whiletherehasbeenlimitedgeneralizable discussionofthe

retentioncharacteristicsofsolutesin SFC,therehasbeenmuch

moreinthecaseofRPLC.Foralkylsilicastationaryphasesusing

acetonitrile,methanol,ortetrahydrofuranwithwaterasthe

elu-ent,Shasbeenempiricallyestimatedat0.25√Mw,whereMw is

themolecularmassofthesolute.InthecaseoftheVal-Aisoform

ofgramicidinwithamolecularmassof1881g/mol,anSvalueof

approximately11 wouldbeexpectedinRPLC;instead,herewe

observe15.7(SupplementaryDataTableS3).Sincethelinear

sol-ventstrengththeoryhasmostlybeenappliedtoreversed-phase

chromatography,itsvalidityinSFChasnotbeenthoroughly

inves-tigated.Glenneetal.recentlyinvestigatedtheretentionofseveral small,unchargedsolutesonaKromasildiolcolumnasafunction

ofthefractionMeOHintheeluent[38].Theypointedoutthatboth thesoluteadsorptiontothestationaryphase aswellasthe co-solventadsorptionneedtobeconsideredtofullyunderstandthe retention[29,38].Furthermore,theydemonstratedthatatalow fractionof co-solventin theeluent,below themaximumofthe MeOHexcessadsorptionisotherm(13v/v%inthatstudy),theLSSis notvalid[29,38].However,atahigherco-solventfraction(asused

inthisstudy),theyfoundthattheLSSmodeldescribesthesolute retentionwell.ThisobservationbyGlenneetal.couldexplainwhy ourexperimentaldesignmodellackedanysignificantquadratic co-solventterms(seeEq.(2)).Thisalsoindicatesthatcautionshouldbe exercisedingeneralizingthetrendspresentedhere,ifthe separa-tionisconductedusingasmallfractionofco-solventintheeluent Onecouldalsonotethatthemaximumoftheco-solventexcess adsorptionisotherm,wheretheLSSmodelbecameacceptablein describingtheretentiontrends,dependsonthetypeofco-solvent (e.g.,MeOH,EtOH,orMeCN)andstationaryphaseusedinthe sepa-ration[39].Forwater,wedonothaveanyadsorptiondataandcan thereforeonlyspeculatethatthewateradsorptiontothe station-aryphaseisstrong.However,inthisexperimentaldesignwedid notobserveanysignificantquadraticwatertermsseeTable2and

Eq.(2);evenifthiscannotbeusedasevidencethereisnowater adsorptionundertheseconditions,thisindicatesthatwithinthis designspacetheeffectofaddingwatertotheeluentfollowsthe LSStheory

TheconclusionthatcanbedrawnfromtheS-valuesisthatour SFCsystemismoresensitivetovariationsintheco-solvent frac-tionthanthecorrespondingRPLCseparationwouldhavebeen.The theoreticalSvaluecouldrepresentahypotheticalsolutewitha molecularmassofapproximately1000gmol−1 intheless sen-sitivesystemandof approximately15,000gmol−1 inthemore sensitivesystem.Furtherstudieswithadiversesetofsolutes, sta-tionaryphases,andeluentswouldgivevaluableinsightintoboth thegeneralretentioncharacteristicsandrobustnessofSFC separa-tions.Ifsmallermoleculestendtobemoresensitivetothestrong eluentinSFCthaninRPLC,usinggradientelutioneven for sep-arationproblemsthatdonotrequiregradientelutiontoachieve reasonableseparationtimeorproductivitymightbebeneficialfrom

arobustnessperspective

Tosummarize,bothgradientandisocraticelutionbecomeless robustiftheseparationsystemunderinvestigationismore sensi-tivetoperturbationsintheparametersunderinvestigation.Using gradientelution,therobustnessincreaseswithincreasinggradient slope

3.5 Practicalimplicationsformethodtransfer

Toputtherobustnesstesting in apractical context,method transferwasconductedforbothanisocraticandagradient sep-arationof gramicidin.In this case,we usedthe centerpoint of theexperimentaldesign BothSFC systems werein their origi-nalfactoryconfigurations,exceptfortheadditionalMSdetector (andpassive flow splitter)on theSFC in Laboratory2 To con-trolthedifferentsystemconfigurations,gramicidininsolutionwas injectedwithoutacolumninbothsystems.Asmalldifferencein injector-detectorvolumewasdetermined,approximately70␮Lin Laboratory1and80␮LinLaboratory2.Thesameidenticalcolumn andidenticalinstrumentalsetconditions(e.g.,back-pressure, tem-perature,gradient,andprograms)wereusedinbothlaboratories TheisocraticchromatogramsfromLaboratories1and2are pre-sentedinFig.7c.Fromthechromatograms,wecanobservearather largedifferencebetweenthe separationsconducted at thetwo laboratories.Toinvestigatetheunderlyingreasonforthelonger retentioninLaboratory2thanLaboratory1,thepressureandmass

Fig 6. Simulated robustness plots based on the experimental gradient system Top row represents separation conducted using a gradient slope of 1% min −1 , center row 7% min−1, and bottom row 13% min−1 The center column comprises robustness plots using the simplified regression model (Eqs 5 and 6) The left- and right-hand columns comprise robustness plots representing theoretical systems less and more sensitive to the co-solvent and water, respectively, by a factor of 2.

Fig 7. Method transfer from Karlstad University (Laboratory 1) to AstraZeneca Gothenburg (Laboratory 2) using the identical column and maintaining the set center-point conditions for the isocratic and gradient methods Contour plot shows the retention factor within the total co-solvent pressure dimension The cross and circle indicate the measured conditions in Laboratories 1 and 2, respectively.

flowsweremeasured.InLaboratory1,thepressureoverthe

col-umnwasmeasuredat162barandthetotalco-solventfractionwas

31.3w/w%,whileinLaboratory2thesamepointwasmeasuredat

159barand29.5w/w%co-solvent.Usingtheisocraticseparation

modelfromtheexperimentaldesign(see section3.3)resultsin

apredictedretentionfactorof6.7±0.7forLaboratory1(Fig.7a,

cross)and9.2±0.7forLaboratory2(Fig.7a,circle).Thismodel predictioncorrespondsverywellwiththeexperimentallyobserved retentionfactorsof7.3and9.3atLaboratories1and2,respectively Gradientseparationwasalsoconducted,andtheresulting chro-matogramsarepresentedinFig.7d.Thedifferencebetweenthe laboratorieswasverysmallinthiscase.Todeterminethe

betweenthelaboratories:Theaveragepressureatthecenterpoint

inthegradientwasmeasuredtobe168barinLaboratory1and

164bar in Laboratory 2 (see Fig.7b).The correspondinginitial

gradientco-solvent(methanol)fractionswere26.6w/w%in

Lab-oratory 1 and 26.7w/w% in Laboratory 2, leading to predicted

(apparent)retentionfactorsof5.6±0.2inbothlaboratories.The

obtainedexperimentalvalueswere5.7and5.6inLaboratories1

and2,respectively

Theexperimentalresultsofthemethodtransfer supportthe

predictionsoftherobustnesscalculations,inwhichgradient

elu-tionispredictedtogiveamorerobustseparationsystemthandoes

isocraticelution.Althoughtheactualconditionsweremore

simi-larinthegradientcase,themainreasonforthemoresuccessful

methodtransferisbelievedtobetheimpactoftheGfactor(Eq.(6))

intheseparationsystem,withseparationsystemshavinghighG

factorslikelybeingmorerobusttodifferencesinsystempressure

andintheactualw/w%ofco-solvent

Theobserveddifferenceinco-solventfractionbetween

Labo-ratories1and2couldhaveseveralorigins,forexample,dueto

differentleakageratesoftheCO2pumpsand/orcheckvalves,

dif-ferentinstrumentconfigurations,orday-to-dayvariationsateither

laboratory[40–42] Theslightly lower pressure atLaboratory 2

couldsimplyhaveresultedfromreducingthetotalflowintothe

back-pressureregulatorbyteeingpartofitintotheMSdetector.It

isworthnotingthattherearenoinstrumentindicationsofthese

differences,astheywereonlyquantifiedusingCFMandpressure

transducersnotpartofthesystem.Thismeansthatfromapractical

perspective,atypicaluserwithoutaccesstopressuretransducers

ormassflowmeterscannotproperlydetectorcompensateforthese

systemdifferencesexceptinanempiricalmanner

It shouldbe noted that both systems in Laboratories 1 and

2wereoperatingwithintheirspecificationsanda

recommenda-tiontousersofmodernanalyticalSFC systemsshouldtherefore

betoalwaysmeasureflow,pressureandcompositionbyexternal

devises,forexamplebyusingthemethodologiesinthisstudy.The

resultscouldbeusedeitherto(I)characterizesystemsindetail(II)

tocalibrateseveraldifferentinstrumentstoperformthesame

per-formanceorto(III)detectifpreventivemaintenanceneedstobe

performed.Tomitigateeffectofdifferentsystemplumbing,stack

configurations,etc.theoperationalconditionscouldbematched

betweenthelaboratoriesaswehavepreviouslydonefor

prepara-tivescale-up[22].However,usinggradientelutionallowsformuch

morerobustoperation,reducingtheneedforcarefulqualification

dependingontherequirementsoftheanalysis

Themethodtransferresultsindicate that,given anidentical

effortinreplicatingtwoseparationsystems,therobustnessofthe

gradientelutionmethodwillleadtomoresuccessfultransferand

shouldbepreferredinseparatinggramicidinusingSFC

4 Conclusions

Therobustnessofpeptideseparationconductedunderisocratic

andgradientconditionsinSFCmodewasinvestigated.Asamodel

system,westudiedthelinearunchargedpentapeptidegramicidin

DseparatedonapH-stablehybridsilicacolumn(Kromasil

SFC-2.5-XT)usinganeluentcontainingCO2,water,andmethanol.The

systemwasfirstcharacterizedusingachemometricDoEapproach

Theexperimentalspacewasthennumericallyexpandedtogain

moregeneralinsight intothesystem.Finally,a gradientandan

isocraticseparationweretransferredtoanotherlaboratorytoput

therobustnesstestinginapracticalcontext

To conduct experimental design, the density of the eluent

(CO2-MeOH-H2O)wasexperimentallydetermined,asnoaccurate

equation-of-statemodelisavailableforthiseluentandweneeded

todeterminetheaveragevolumetricflowrateateachdesignpoint

Wefoundthat Coriolismassflowmeterscouldaccurately mea-surethedensity.Wealsoconcludedthatworkingwithahigh-mass fractionofmethanolandwaterasco-solventsresultedinsmall vari-ationsindensityover alargeareaofpressureandtemperature, inherentlymakingSFCmorerobust

FromtheDoE,wefoundthatthetotalfractionofco-solventin theeluentandthewaterfractionintheco-solventwerethemost importantfactorscontrollingtheretention.Themeasured sensitiv-itywashigherthantheRPLCvaluesforsimilarseparationsreported

intheliterature.Wealsofoundthatingradientelution,the sepa-rationisatleastthreetimesmorerobusttoperturbationsthanin isocraticelution

InspiredbytheDoEmodel,weinvestigatedsystemsthatare more and less sensitive tochanges in the eluent composite as wellasgradientelutionconductedatdifferentgradientslopes.We foundthatbothgradientandisocraticelutionsbecomelessrobust formoresensitivesystems.Usinggradientelution,therobustness increaseswithincreasinggradientslope

Finally,themethodsweretransferredtoanotherlaboratory.The resultsoftheisocraticmethoddifferedgreatlybetweenthe labora-tories,themainreasonsforthisbeingdifferencesinpressureand

inthetotalco-solventfractionbetweenthesystems.Thisdeviation couldbeexplainedusingtheDoEmodel.Forthegradient separa-tion,thetransferwassuccessful.Theresultsclearlyindicatethat gradientelutionresultedinaconsiderablymorerobustseparation system

Acknowledgements

Thisworkwassupportedby(i)theSwedishKnowledge Foun-dation for the KKS SYNERGY project 2016 “BIO-QC: Quality ControlandPurificationforNewBiologicalDrugs”(grantnumber 20170059),by(ii)theSwedishResearchCouncil(VR)fortheproject

“FundamentalStudiesonMolecularInteractionsaimedat Prepar-ativeSeparationsandBiospecificMeasurements”(grantnumber 2015–04627),by(iii)theÅForskFoundationfortheproject “Qual-itycontrolofnextgenerationbiologicalbasedmedicines”(grant number17/500)andby(iv)thegrant2015/18/M/ST8/00349from theNationalScienceCentre,Poland)

Appendix A Supplementary data

Supplementarymaterialrelatedtothisarticlecanbefound,in theonlineversion, atdoi:https://doi.org/10.1016/j.chroma.2018 07.029

References

[1] U Gottschalk, K Brorson, A.A Shukla, The need for innovation in biomanufacturing, Nat Biotechnol 30 (2012) 489–492, http://dx.doi.org/10 1038/nbt.2263

[2] E Lesellier, C West, The many faces of packed column supercritical fluid chromatography – a critical review, Ed Choice IX 1382 (2015) 2–46, http://dx doi.org/10.1016/j.chroma.2014.12.083

[3] G Guiochon, A Tarafder, Fundamental challenges and opportunities for preparative supercritical fluid chromatography, J Chromatogr A 1218 (2011) 1037–1114, http://dx.doi.org/10.1016/j.chroma.2010.12.047

[4] J Zheng, J.D Pinkston, P.H Zoutendam, L.T Taylor, Feasibility of Supercritical Fluid Chromatography/Mass Spectrometry of Polypeptides with Up to 40-Mers, Anal Chem 78 (2006) 1535–1545, http://dx.doi.org/10.1021/ ac052025s

[5] D Tognarelli, A Tsukamoto, J Caldwell, W Caldwell, Rapid peptide separation by supercritical fluid chromatography, Bioanalysis 2 (2010) 5–7,

http://dx.doi.org/10.4155/bio.09.165 [6] B Bola ˜ nos, M Greig, M Ventura, W Farrell, C.M Aurigemma, H Li, T.L Quenzer, K Tivel, J.M.R Bylund, P Tran, C Pham, D Phillipson, SFC/MS in drug discovery at Pfizer, La jolla, Int J Mass Spectrom 238 (2004) 85–97, http://dx doi.org/10.1016/j.ijms.2003.11.021