Programmable droplet-based microfluidic serial dilutor

Programmable

droplet-based

micro

fluidic

serial

dilutor

Hoon

Suk

Rho

a,b,c

,

Yoonsun

Yang

d

,

Leon

W.M.M.

Terstappen

d

,

Han

Gardeniers

b

,

Séverine

Le

Gac

c,

**

,1

_,

_Pamela

_Habibovic

a,

_*

,1

a_{DepartmentofInstructiveBiomaterialsEngineering,MERLNInstituteforTechnology-InspiredRegenerativeMedicine,MaastrichtUniversity,TheNetherlands} b_{MesoscaleChemicalSystemsGroup,MESA+Institute,UniversityofTwente,TheNetherlands}

c

AppliedMicrofluidicsforBioEngineeringResearchGroup,MESA+Institute,TechMedInstitute,UniversityofTwente,TheNetherlands

d

MedicalCellBioPhysicsGroup,TechMedInstitute,UniversityofTwente,TheNetherlands

ARTICLE INFO

Articlehistory: Received28June2020

Receivedinrevisedform1August2020 Accepted2August2020

Availableonline11August2020

Keywords: Microfluidics Droplet Serialdilution Microvalve ABSTRACT

Aprogrammabledroplet-basedmicrofluidicserialdilutorplatformispresented,whichiscapableof generatingaseriesofdropletswiththescalablestepwiseconcentrationgradientofasample.Sequential dilution of a target molecule was automatically performed in sub-nanoliter scale droplets by synchronizingamicrofluidicperistalticmixerandavalve-assisteddropletgenerator.Thevolumeof dropletsdispensedfromthemixerwascontrolledbymicrovalveoperation,whichenabledtotunethe dilutionwithvariousdilutionfactors.Afterevaluationofthemixerefficiencyandcalibrationofthe dropletsizeatdifferentvalveoperating conditions,serialdilutionsofrhodamineB isothiocyanate-dextranwasdemonstrated,inanautomatedmanner,atthreedifferentdilutionfactors.Specifically,the effectoftherhodamineBisothiocyanate-dextranconcentrationandtemperatureonvariationsofthe fluorescentintensitywasquantified.Thisprogrammablemicrofluidicdropletserialdilutorwillopennew avenues,ananalyticaltool,toevaluatecomplexchemicalandbiochemicalreactions,especiallywhen limitedsamplevolumeisavailable,forexample,attheearlystageofdrugdiscoveryandbiochemical processdeveloping.

©2020TheAuthors.PublishedbyElsevierB.V.onbehalfofTheKoreanSocietyofIndustrialand EngineeringChemistry.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons. org/licenses/by/4.0/).

Introduction

Serialdilution,aseriesofsequentialdilutionsofasample,isone of themostcommon and necessaryprocesses in chemical and biologicallaboratoriesforsamplepreparationand experimenta-tion.Furthermore,serialdilution is afundamental operationto evaluateever-increasingsamplenumbersinvariousconcentration rangestoidentifythedynamicinteractionoftargetmoleculeswith (bio)chemical libraries [1,2]. The universal technique for serial dilutionreliesonpipettinginasetofmicrotubesorinamicro-well plate,which howeverrequiresarelativelylargesamplevolume andextensivemanualhandlingthatmightresultincompounding errors. Liquid handling automation, using dedicated robots, especiallyincombinationwithtestsminiaturization,automation, andcomputationaldataprocessing,isanexcellentalternativefor

high-throughputscreening(HTS)forbiochemicaland biopharma-ceuticalapplications[3].Althoughliquidhandlingrobotsare well-establishedfortheseHTSpurposeshigh-throughputscreeningin thepharmaceuticalandbiotechnologyindustries,thesesystems are costly and require considerable space for installation and operation. Advances in microfluidic technology have proposed moreintegratedapproachesforhandlingcomplexfluidflowsand producing concentration gradients in miniaturized platforms

[4,5,14–23,6–13].Microfluidicplatformsofferavarietyofbenefits, suchashavingasmallfootprint,lowsampleconsumption,shorter analysistime,andlowerassaycost.Especially,parallelized lab-on-a-chip devicescoupledwithautomatedoperationshavea great potential for performing sequential biological and chemical reactions,whichhasrelevancenotonlyforexperimentalscience inthelaboratorybutalsoforindustrialapplications.

Varying the reagent concentrations using microfluidics is essential for performing fast analytical screens which requires multiple parallel reactions, e.g., chemotaxis [24,25], haptotaxis

[26],cytotoxicity[7,9,27],immunoassay[6],nucleicacidpuri fica-tion[12],andenzymaticreactions[14].However,thisoperationis challenginginamicrofluidicformatbecausemixingreliesonlyon diffusion.Toovercomeallthesehandlingandmixinglimitations, severalmicrofluidicdeviceshavebeenproposedforserialdilution

* Correspondingauthorat:MaastrichtUniversity,Universiteitssingel40,Room C3.577,6229ERMaastricht,TheNetherlands.

** Corresponding authorat: Universityof Twente, Building Zuidhorst,Room ZH135,7500AEEnschede,TheNetherlands.

E-mailaddresses:s.legac@utwente.nl(S.LeGac),

p.habibovic@maastrichtuniversity.nl(P.Habibovic).

1

Equalcontributions.

https://doi.org/10.1016/j.jiec.2020.08.004

1226-086X/©2020TheAuthors.PublishedbyElsevierB.V.onbehalfofTheKoreanSocietyofIndustrialandEngineeringChemistry.ThisisanopenaccessarticleundertheCC BYlicense(http://creativecommons.org/licenses/by/4.0/).

ContentslistsavailableatScienceDirect

Journal

of

Industrial

and

Engineering

Chemistry

toobtain concentration gradientsof a sample, using networks of continuous flows, large-scale integration with microvalves

[10–16],andmicrodroplets[17–23],asalsoreviewedbyKennan etal.[28].

ByinterconnectingmultiplemicrofluidicT-mixersin continu-ousflowbranches,aChristmastree-shapedchannelnetworkwas developedforgeneratinglinearsequentialconcentrationgradients

[4,5].Toenhancemixingperformance onareduced footprint,a varietyof mixers and designs have been integrated, such as a herringbone mixer [6], 3D fluidic channel connections [7], a serpentineTesla mixer [27],and high volumemixing channels

[8,9].Suchmicro_fluidicplatformsallsupporttheformationofa linear, complex, and combinatorial concentration gradient in continuousfluid flows,which is beneficialfor analytical assays wherereagent perfusionis necessary, e.g.,cell-basedscreening. However, high-throughput screening for quantitative analysis, especiallywhenlimitedresourcesareavailable,remains challeng-ingbecausecreatingaconcentrationgradientinacontinuousflow consumesarelativelylargeamountofsample.

Microvalvetechnologiesprovideaccuratecontrolofmultiple fluidflowsandactiveperistalticmixingofsolutionsinlarge-scale systems with integrated reactors [10,11]. Sequential and/or combinatorial concentration gradients have been generated by metering,combining,andmixingmultiplereagentsinparallelized nanoliter reactors [12–15]. Also, the in-line arrangement of multipleperistalticmixersenablestheserialdilutionofatarget reagent by mixing stored solutions in each dilution stage sequentially [16]. Although valve-assisted devices have proven tobeparticularlypromisingforincreasingtheintegrationpotential forreactor-basedscreening,thecombinationofreagentsandthe resultingconcentrationgradientshavesofarbeenpredetermined bythedevicedesign,e.g.,thesizeofreactorsand/orthechannel configuration.

Droplet-basedmicrofluidicshasprovidedanewparadigmto conduct reactions in a rapid and robust manner, with great flexibility in the size of the reactors, down to extremely tiny volumes[17–22].In this discreteconfiguration,ina continuous flownetwork,targetsubstancesinsolutionsareinjectedintoan immisciblecarrier fluid to formnano-to-picoliter droplets, and concentration gradients were for instance obtained by mixing solutions before droplet formation [18], by merging already formed droplets [19,20], by injecting diluent droplets into an existing sample droplet [17], or by passing a sample droplet throughdiluentdroplets[21]indroplettrappingunits.However, inpractice,itisdifficulttoobtainfastandstablechangesin

flow-ratesofmultiplefluids,evenwhenusingprecisionsyringe-pumps. Toaddressthischallenge,microvalveshavebeenintegratedwith droplet-microfluidic devices for precise and programmable dropletmanipulation[22,23].Microvalve-assisteddroplet gener-atorsprovidemoreflexibilitybymechanicalcuttingofadispensed flowandinjectingnewdropletsintoformeddropletstoeventually createaconcentrationgradient.However,a disadvantageofthe droplet-merging-based dilution approach is to keep the final volume of the droplets constant, which requires continuous multipledropletsizecontroland synchronizedmergingateach dropletgeneratingprocessarerequired.

Here,wereportanautomateddroplet-basedmicro_fluidicserial dilutor,inwhichaseriesofdropletsisgeneratedwithexcellent controlontheirsize,tocreateaconcentrationgradientofatarget moleculewithflexibledilutionscales.Themicrofluidicplatform comprisesaperistalticmixerandadropletgeneratortodilutethe targetreagentinsub-nanoliterwaterdropletsinanoilcarrierfluid. Theperformanceoftheperistalticmixingunitwasfirstoptimized byvaryingtheoperatingconditionsofthemicrovalves.Also,the dropletsizewascalibratedbychangingtheappliedpressuresfor thewaterandoilphasesaswellasthedispensingtimetoprogram thedilutionfactoroftheserialdilution.Bysynchronizingthemixer and the droplet generator, a microfluidic serial dilution of rhodamine B isothiocyanate-dextran (RD) was automated with dilutionfactorsof7.69,5.32,and2.70,andtheinfluenceofboththe RDconcentrationandthetemperatureonthefluorescentintensity was evaluated. The programmable droplet-based microfluidic serialdilutormightbeusefulforperformingcomplexchemicaland biochemicalreactionsinexperimentalsciencesandengineering, particularly where tiny sample consumption is preferred, e.g., biochemical process development and drug discovery and screening.

Materialsandmethods

In-lineserialdilutionwithflexibledilutionfactors

A conventional serial dilution method follows a sequential dilutionprocedurewhereeachstepincludesthepreparationofa diluent, thetransfer of analiquot of stocksolution, and actual mixing (Fig. 1A). In contrast to this stepwise process, the microfluidicdroplet-based serialdilutionwasconceptualized to conduct all the required procedures in-line with on-demand dilution factors. The design of the microfluidic droplet serial dilutorincludedamixer,aclosedreactorthatwasinitiallyfully

occupiedwitha sample solution,connectedtoan inlet andan outlet(Fig.1B).Theinletandoutletareequippedwithvalvesthat wereinterconnectedtosynchronizetheopeningandclosingtimes. Hence,whenacertainvolumeofadiluentwasintroducedintothe mixingreservoir,thesamevolumeofthesolutionwasdispensed fromthemixer,consequentlyresultinginasolutionconcentration decreaseinthemixer.Thegoverningequationforcalculatingthe concentrationchangeofthesolutioninthemixerafteraddingthe diluent(Cn+1)is:

Cnþ1¼VinCinþðV_V0VoutÞCn

0 ð1Þ

where Cn, Cin, V0, Vin, and Vout are the concentrationof the

solutioninthemixer,theconcentrationofthediluentintroduced intothemixer,thevolumeofthesolutioninthemixer,thevolume ofthediluentaddedintothemixer,andthevolumeofthesolution dispensedfromthemixer,respectively.Whentheinputandoutput volumesarethesame,andtheconcentrationofthediluentiszero,

theequationcanbesimplifiedas: Cnþ1¼V0_VVin

0

Cn ð2Þ

Hence,thedilutionfactorisdeterminedbythevolumeofthe introduceddiluentatafixedmixersize.Forexample,logarithmic, half-logarithmic, and quarter-logarithmic dilutions could be obtainedbyadding101,3.161,and1.781oftheinitialvolume ofthesolutioninthemixer(V0).

Designofthemicrofluidicdroplet-basedserialdilutor

The device comprised a microfluidic mixer, a valve-assisted droplet generator,and two dropletincubation stages. The chip architectureincludedtwolayers,atopfluidicchannellayeranda bottom valve-control channel layer. The designs of channel networks,solutioninletsandoutlets,andcontrolportsareshown inFig.2A.Theflowchannelsforloadingacarrierfluid(greycolor),

Fig.2.Thedesignandoperationofthemicrofluidicdroplet-basedserialdilutor.(A)CADdesignofthechip.(B)Microscopicimageofafabricateddevice.(C)Operationofthe device.On-chipserialdilutionwasprocessedbyrepeatingthreesteps,(a,d)pushingthesolutioninthemixingunitwithadiluent,(b,e)dropletformation,and(c,f) peristalticmixingofthediluentandthesamplesolution.Photoswereextractedfromarecordedvideo.Scalebars,300mm.(Forinterpretationofthereferencestocolourin thetext,thereaderisreferredtothewebversionofthisarticle.)

adiluent(lightgreycolor),andasamplesolution(bluecolor)were connectedtotheinletandoutletinthefluidiclayer.Thechannels tocontrol theshut-off(red color) and mixing valves (magenta color)wereconnectedtodedicatedportsinthecontrollayer.The mixingunitconnectedtothesample-,diluent-,andcarrierfluid channelsandshut-offvalveswerelocatedatthechanneljunctions toopenandclosetheconnections(Fig.2B).Fig.2Cillustratesthe processflowofrepeated droplet-basedserialdilutions.Initially, thesamplesolutionwasloadedintothemixerbyopeningavalve forthereagentchannel,whilethevalvesnearthediluentchannel andoilchannelwereclosed.Then,allvalveswereclosedtoisolate themixer,andadropletwasdispensedintothecarrier_fluidflowby openingthevalvesnearthediluentandoilchannels(Fig.2C(a)). Thesizeofthedropletwas determinedbycontrollingthevalve openingtime.Afterformingthedroplet(Fig.2C(b)),threemixing valves were sequentially actuated to generate a fluid flow for enhancedmixingofthesamplesolutionanddiluent(Fig.2C(c)). Byrepeating theprocesses ofdroplet formation and peristaltic mixing(Fig.2C(d)–(f)),aserialdilutionofthesamplesolutionwas achieved in a series of homogeneously-sized droplets, and all procedureswereautomatedthroughproperprogrammingofthe valveoperatingsequences.SupplementaryVideoS1presentsthe automated processes of the microfluidic droplet-based serial dilution.

Devicefabrication

The microfluidic chip was fabricated by multilayer soft lithographyusingpolydimethylsiloxane(PDMS)[10,29],following themodifiedprotocolsfromourpreviouswork[14,30–32]. Deviceoperation

A pneumatic control set-up was used for operating the microfluidic serial dilutor. 3/2-way solenoid valves, precision pressure regulators, and EasyPort USB digital I/O controller wereobtained from Festo(FestoB.V., Delft, TheNetherlands) and combined by applying compressed nitrogen gas at a pressureof 1.3bar inthe controlchannels.Thecontrol setup wasautomatedbyanin-housebuiltLabVIEW(National Instru-ments Co., Austin, USA) program to actuate the valves sequentiallywithamaximumresponsetimeof5ms.Solutions were introduced into the fluidic channels by applying com-pressednitrogen gasonthe backsideof solutionsthrough 3D printedplugs.Thecalibrationcoefficientofthepressure-driven flowforwaterwas0.033

m

Lmin1mbar1inourpreviousstudy

[33].Theappliedpressureforthecarrierfluidwasvariedfrom 160 to 260mbar to evaluate its impact on the droplet size controlwhileaconstantpressureof200mbarwasappliedfor loadingtheaqueoussolution.

Dataprocessing

A stereomicroscope(MoticSMZ171-TLED, LabAgency Bene-lux B.V., Dordrecht, The Netherlands) and a CMOS camera (Moticam3.0)wereusedformonitoringmixinginthemixerand thedroplet sizecontrol inthe droplet generator.An inverted fluorescent microscope (Olympus IX73, Olympus Netherlands BV,Leiderdorp,The Netherlands)equippedwithan automatic XY-stage(99S000,LudlElectronicProductsLtd.,NY,USA),anda digital camera (ORCA-ER, Hamamatsu Photonics Deutschland GmbH,Herrsching,Germany)wasautomatedusinganin-house builtLabVIEW(NationalInstrumentsCo.,Austin,USA)program to acquire fluorescence images of RD droplets. All acquired imageswereanalyzedbyImageJsoftware(http://rsb.info.nih. gov/ij/).

Devicevalidation(peristalticmixinganddropletgeneration) A food dye solution filtered with a 0.2-

m

m syringe filter (WhatmanPLC,Sigma-Aldrich,Zwijndrecht,TheNetherlands)and deionizedwaterfromMilli-Qfiltrationsystem(MilliporeCo.)were usedforthecharacterizationofperistalticmixinginthedevice. Mineraloilcontaining1.5%(w/w)Span80(allfromSigma-Aldrich ChemieBV,Zwijndrecht,TheNetherlands)wasusedasthecarrier fluidfortheformationofthewater-in-oil(w/o)droplets. Theeffectofconcentrationandtemperatureonthefluorescence intensity

1000mgL1rhodamineBisothiocyanate-dextran(RD,average moleculeweight10,000,Sigma-AldrichChemieBV,Zwijndrecht, TheNetherlands)preparedin Milli-Qwater(Millipore Co.)was usedasasamplesolutionforcharacterizingtheinfluenceofthe temperatureonthefluorescencesignal.Thedextranconjugatewas selectedtopreventthepenetrationofrhodamineBmoleculesinto PDMSchannels[34].TheRDsolutionwasintroducedinthemixing unitinthechipanddilutedwithMilli-Qwatertoobtainaserial dilutioninaseriesofdroplets.Thedropletswerecollectedinthe incubationstages,andtheRDfluorescentsignalofthedropletswas quantified by N 2.1 filter cube (excitation: BP 515–560nm; emission: LP 590nm).An indium tin oxide (ITO)heater and a controller were obtainedfromCell MicroControls (Norfolk,VA, USA) and calibrated tovary thetemperature in theincubation stagesinthemicrofluidicdevice[32].Notethatthetemperature was changed after performing a serial dilution in a series of droplets, so potential changes in solution viscosity or PDMS elasticity due to the temperature change, which could affect dropletsizeanddilution,isavoided.Furthermore,thefluorescent intensitywas analyzedinthewholeareaofthedroplet,which meansthatthetotalamountoffluorescentmoleculesisrecorded. Hence,anydropletsizechangeduetotemperaturewillnotaffect theobtainedfluorescentintensitiesoftheRDdroplets.

Resultsanddiscussion

Validationofthemixingcapabilityoftheperistalticmixer

Microfluidicperistalticmixing,oneoftheessentialfunctionsof themicrofluidicdroplet-basedserialdilutor,wasfirstevaluatedto optimizethemixingperformanceaswellastosynchronize the mixing and droplet generation processes. Fig. 3A depicts the processflowofperistalticmixinginthedevice.First,afilteredblue fooddyesolutionwasloadedintothemixerthroughthesample channelandisolatedbyclosingallvalvesnearthemixer(Fig.3A (a)). Then, Milli-Q water was introduced by simultaneously opening two valves near the diluent and the carrier fluidic channels(Fig.3A(b)).Sinceaconstantpressureof200mbarand 220mbarwasappliedonthebacksideofthewaterandoilphases, respectively,thevolumeofMilli-Qwaterinjectedintothemixer wasdeterminedbycontrollingtheopeningtimeofthevalves.By loadingMilli-Qwaterwithavalveopeningtimeof200ms,halfof themixervolumewasfilledwithMilli-Qwater.Afterloading Milli-Qwater, the solutionswere mixed,and theaverage brightness valueofthesolutioninthetotalareaofthemixerwasmonitored overtime(Fig.3A(c)and(d)).Fig.3B(left)presentsthebrightness valuechangein themixerwithout theactuationofthemixing valves.Theaveragebrightnessvaluedecreasedwiththemixing time until a constant value was reached when mixing was completed.Usingdiffusiononly,themixingoftwosolutionswas completedafter10min.

Toenhancethemixingperformance,threevalveswereactuated withthesequence(100),(110),(010),(011),(001),and(101),where

0meansthatthevalveisopen,and1thatitisclosed.Intheclosed loop-shapedmixer,theperistalticpumpinggeneratedtherotation offluidsandstretchedtheparabolicprofileofthePoiseuilleflow

[11].Thelongstretchedinterfacebetweenthefluidsenabledthe enhanceddiffusionofmoleculesinthefluids.Apreviousstudyona microfluidicperistaltic mixerreportedthatalargevalvearea,a small distance between valves, and a fast operating frequency increasedmixing efficiency [35]. Hence, the microfluidic serial dilutor included mixing three valves with a surface area of 100150

m

m2andaninter-valvedistanceof75

m

m,whichwere identified as the optimal dimensions for this particular device design and fabrication. Fig. 3B (right) illustrates the complete mixingtimeatvariousvalveoperatingfrequenciesrangingfrom1 to 40Hz. The complete mixing time was found to decrease exponentiallywiththeincreasingvalveoperationfrequency,and solutionsweremixedin2satoperatingfrequenciesabove5Hz. Following to this mixer characterization, a valve operating frequency of 20Hz and a mixing time of 2s were used for programminganin-housebuiltautomatedoperationsoftwarefor theserialdilutionchip.Mixingsolutionsinthedeviceatvarious valve operating frequencies are presented in Video S2 in supplementarymaterials.

Controlonthedropletsize

The dilution factor in the microfluidic droplet-based serial dilutionisdeterminedbythevolumeofthediluentloadedintothe mixer,whichisequaltothevolumeofdispensedsamplesolution volume.Hence,thecapabilityofthedevicetoformmonodispersed dropletswithawell-de_finedsizeisessentialtoachieveaccurate serialdilutionwithtunabledilutionfactorsofthedilutionfactors. The size of the droplets formed by a valve-assisted droplet generatortypicallydependsontheflow-rateratiooftheaqueous phase and the carrier fluid as well as the dispensing (valve opening)time[22].Tocalibratethedropletsizeasafunctionofthe deviceoperatingconditions, dropletsweregeneratedatvarious pressuresappliedforthewaterandoilphasesaswellasdifferent dispensingtimes,andanalyzed.Thefilteredfooddyesolutionand themineraloilsupplementedwith1.5%(w/w)Span80wereused astheaqueousphaseandcarrierfluid,respectively,forgenerating w/o droplets.The dye solutionwasintroduced intothediluent channelanddispensedintotheoilphasebyopeningtwovalves, onelocatedbetweenthediluentchannelandthemixer,andthe otheronebetweenthemixerandtheoilchannel,concurrently.The valveopeningtime for dispensingthewater phase intotheoil

Fig.3. Mixingsolutionsintheperistalticmixer.(A)Sequenceillustratingthemixingofsolutions.(B)MixingofafooddyesolutionandMill-Qwateronlybydiffusion(left)and usingthemixingvalvesactuatedatvariousoperatingfrequencies(right).Scalebars,200mm.

phase was varied from 40 to 200ms. A constant pressure of 200mbar was appliedfor loadingthe oil phase, while various pressuresranging from 160 to 260mbar were applied for the aqueousphase tovarythepressureratiosfor thewaterand oil phasesfrom 0.8 to1.3. The dropletformation process atthese variouspressureratios,whileusingaconstantdispensingtimeof 100ms,ispresentedinFig.4A:alinearincreaseinthedropletsize wasobservedwhenincreasingtheappliedpressureforloadingthe waterphase.Next,whilemaintainingthesameappliedpressures of 200mbar for both the aqueous and carrier fluid flows, the dispensingtimewasincreased,whichasshowninFig.4Bresultsin theproductionoflargerdroplets.Forcalibratingthevolumeofthe formeddroplets accordingto variousoperating conditions, the dropletswerecollectedfromtheoutletofthedeviceandplacedon

acavityslide(PaulMarienfeldGmbH&Co.KG,Lauda-Königshofen, Germany)toensurethatthedropletsretainedasphericalshape. The measured volume of the droplets was summarized in the graphsinFig.4C.Thedropletvolumesrangedfrom0.190.01nlto 3.820.02nlatdifferentoperatingconditions.Theinfluenceof boththeappliedpressureratiosforthewaterandoilphasesand the dispensing time on the droplet volume is summarized in

Fig. 4C. The linearity of thefitting lines and the coefficient of determination(R2_{0.986,n=20)highlighttherobustnessofour}

device to produce monodispersed droplets. Since the mixer volumewasmaintainedconstantandequalto5nl,thedilution factorinthedevicecouldbeprogrammedon-demandbyselecting theoperatingconditions,e.g.,appliedpressuresforthedifferent fluid flowsand thedispensingtime, basedonthesecalibration

Fig.4. Controllingthedropletsizebyvaryingtheoperatingconditions.Dropletsofdifferentsizeswereformedby(A)varyingtheratiosofthepressuresappliedforthewater andoilphasesataconstantdispensingtime(100ms),and(B)varyingthedispensingtimeataconstantpressureratioforthewaterandoilphases(1.0).Scalebars,200mm.(C) Relationshipbetweenthedropletvolumeandthepressureratio(left)andthedispensingtime(right)(n=20).

curves.Thedropletgenerationprocessatvariousdispensingtimes isshowninVideoS3insupplementarymaterials.

Serialdilutionofrhodamine-dextran(RD):effectofthetemperature Concentration gradients of RD were created in a series of dropletsasa proofofconcepttodemonstrateon-demandserial dilution in the device. Before starting the microfluidic serial dilution, the fluorescence intensity of the RD droplets was calibratedasafunctionoftheirconcentration.RDsolutionswith various concentrations, 1000mgL1, 500mgL1, 250mgL1, 125mgL1, 62.5mgL1, 31.25mgL1, and 15.625mgL1, were preparedinMill-Qwateranddispensedfromthediluentchannel intothecarrierfluid(mineraloilcontaining1.5%(w/w)Span80)in theoilchanneltoformw/odroplets,andthefluorescenceintensity of the droplets was quantified. As shown in Fig. 5A, the RD fluorescenceintensityinthedropletsvarieslinearlyasafunction oftheRDconcentrationinthetestedrange(n=10).

Forperformingdroplet-basedmicrofluidicRDserialdilutions, 1000mgL1oftheRDsolutionwasloadedintothemixer,followed byMill-Qwater,whichpushedtheRDsolutiontoformRDdroplets inthecarrierfluid.Theappliedpressureforboththewaterandoil phases was 200mbar, and three different dispensing times, 66.7ms,100ms,and200ms,weretestedforgeneratingdroplets havingvolumesof0.650.01nl,0.940.01nl,and1.850.02nl, respectively(n=3).Fortheseconditions,thedilutionfactors,as derivedfromthemeasureddropletvolume,were7.69,5.32,and 2.70,respectively.TherelationshipbetweentheRDdropletindex andtheRDconcentrationusing thedifferentdilutionfactorsis presentedusinglinear(Fig.5B) andlogarithmic(Fig.5C) scales

(n=3).TheinitialconcentrationoftheRDsolution,1000mgL1, wasloweredabout100timesover35,24,and11dropletsby 7.69-fold,5.32-fold,and2.70-foldserialdilutions.

To characterize the effect of the temperature, a series of dropletswithRDconcentrationsvaryingfrom1000to19mgL1, generated by a 5.32-fold droplet-based serial dilution, was collected in the incubation stage in the device. Thereafter,the temperaturewas increasedfrom25Cupto50Cwithstepsof 2.5_{Canddecreasedagainto25}_{C.Ateachtemperature,theRD} fluorescenceintensityofthedropletswasmeasuredandcompared with theinitial intensity. Thefluorescence imagesof theseRD dropletswithvariousconcentrationsatdifferenttemperaturesare shown in Video S4 in supplementary materials, and Fig. 6

illustratestheeffectofthetemperatureandtheRDconcentration on the droplet fluorescence intensity. The RD fluorescence intensity was found to decrease as the temperatureincreased, and this decreasewasmorepronounced forhigherRD concen-trations, reaching a plateau of 1.96 %/C. The concentration-dependentdecreaseratemighthaveresultedfromthe reabsorp-tion of emitted light or collisional quenching behavior of fluorescence moleculesathigher densities [36]. Thismaximum decreaserate isin closeagreementwithvaluesreportedinthe literature of1.90%/C, whichwasmeasuredfor rhodamineBin water usinga spectrofluorometer equippedwithatemperature control module[37].The reliablefluorometric measurement of moleculesindropletsassociatedwiththefastheattransferinsuch smallvolumes,providingexquisitecontrolonthetemperatureisof greatinteresttomonitorthetemperature-dependentkineticsof reactionswithfluorescencemarkersforbiotechnologicalstudies, e.g.,proteinaggregation[32,38],andenzymekinetics[14,23].

Fig.5.Droplet-basedserialdilutionofRDwithvariousdilutionfactors.(A)RelationshipbetweentheconcentrationandfluorescenceintensityofRD(n=10).(B)Droplet indexandRDconcentrationsofseriallydilutedRDdropletsplottedonalinearscale(n=3).(C)LogarithmicregressionofthedropletindexandtheRDconcentration.Dashed linesrepresentcalculatedvalues.1000mgL1ofRDsolutionwasdilutedinaseriesofdropletsinthedeviceataconstantappliedpressureratio(Pwater/Poil)of1.0andusing

Conclusion

In conclusion, we have developed a programmable droplet-basedmicrofluidicserialdilutorforcreatingon-demand concen-trationgradientsofasamplesolution.Themicro_fluidicperistaltic mixer and the valve-assisted droplet generator were first calibratedandsynchronizedtoachievefastmixing andreliably controlthe droplet size for droplet-based serial dilution. Next, serialdilutionsofrhodamineBisothiocyanate-dextran(RD)was demonstratedatvariousdilutionfactors.Theautomated program-mableserialdilutioninaseriesofnanoliterdropletswillopennew avenuestoevaluatechemicalandbiochemicalreactionsin high-throughputscreeningprocesses.

Conflictofinterest

Theauthorshavenoconflicttodeclare. CRediTauthorshipcontributionstatement

Hoon Suk Rho: Conceptualization, Methodology, Software, Validation,Datacuration,Formalanalysis,Investigation, Visuali-zation, Writing - original draft, Writing - review & editing. YoonsunYang:Conceptualization,Datacuration,Formalanalysis, Investigation,Writing-review&editing.LeonW.M.M. Terstap-pen: Funding acquisition, Writing - review & editing. Han Gardeniers:Methodology,Writing -review&editing. Séverine LeGac:Methodology,Writing-originaldraft,Writing-review& editing. Pamela Habibovic: Methodology,Supervision, Funding acquisition,Writing-originaldraft,Writing-review&editing. Acknowledgments

Thisworkwas_financiallysupportedbythePerspectiefProgram CancerID(sub-project#14193),InnovationalResearchIncentives SchemeVidi(#15604)oftheNetherlandsOrganizationforScientific Research (NWO), The Netherlands, and the Dutch Province of Limburg,TheNetherlands.PHgratefullyacknowledgesthe Gravita-tionProgram‘Materials-DrivenRegeneration’,fundedbytheNWO. AppendixA.Supplementarydata

Supplementarymaterialrelatedtothisarticlecanbefound,inthe onlineversion,atdoi:https://doi.org/10.1016/j.jiec.2020.08.004.

References

[1]T.R.Johnson,C.L.Case,LaboratoryExperimentsin Microbiology,Pearson/ BenjaminCummings,2004.

[2]K.R.Aneja,ExperimentsinMicrobiology,PlantPathologyandBiotechnology, NewAgeInternational,2007.

[3]L.M.Mayr,D.Bojanic,Curr.Opin.Pharmacol.9(5)(2009)580,doi:http://dx. doi.org/10.1016/j.coph.2009.08.004.

[4]N.L.Jeon,S.K.W.Dertinger,D.T.Chiu,I.S.Choi,A.D.Stroock,G.M.Whitesides, Langmuir16(22)(2000)8311,doi:http://dx.doi.org/10.1021/la000600b. [5]S.K.W.Dertinger,D.T.Chiu,NooLiJeon,G.M.Whitesides,Anal.Chem.73(6)

(2001)1240,doi:http://dx.doi.org/10.1021/ac001132d.

[6]X.Jiang,J.M.K.Ng,A.D.Stroock,S.K.W.Dertinger,G.M.Whitesides,J.Am.Chem. Soc.125(18)(2003)5294,doi:http://dx.doi.org/10.1021/ja034566+. [7]G.M.Walker,N.Monteiro-Riviere,J.Rouse,A.T.O’Neill,LabChip7(2)(2007)

226,doi:http://dx.doi.org/10.1039/b608990a.

[8]K.Hattori,S.Sugiura,T.Kanamori,LabChip9(12)(2009)1763,doi:http://dx. doi.org/10.1039/b816995k.

[9]S.Sugiura,K.Hattori,T.Kanamori,Anal.Chem.82(19)(2010)8278,doi:http:// dx.doi.org/10.1021/ac1017666.

[10]M.A.Unger,H.P.Chou,T.Thorsen,A.Scherer,S.R.Quake,Science(80-.)288 (5463)(2000)113,doi:http://dx.doi.org/10.1126/science.288.5463.113. [11]H.P.Chou,M.A.Unger,S.R.Quake,Biomed.Microdevices3(4)(2001)323,doi:

http://dx.doi.org/10.1023/A:1012412916446.

[12]J.W.Hong,V.Studer,G.Hang,W.F.Anderson,S.R.Quake,Nat.Biotechnol.22(4) (2004)435,doi:http://dx.doi.org/10.1038/nbt951.

[13]H.S.Rho,A.T.Hanke,M.Ottens,H.J.G.E.Gardeniers,Electrophoresis39(8) (2018)1031,doi:http://dx.doi.org/10.1002/elps.201700341.

[14]H.S.Rho,A.T.Hanke,M.Ottens,H.Gardeniers,PLoSOne11(4)(2016),doi: http://dx.doi.org/10.1371/journal.pone.0153437e0153437.

[15]Y.Yang,S.LeGac,L.W.M.M.Terstappen,H.S.Rho,Electrophoresis39(3)(2018) 548,doi:http://dx.doi.org/10.1002/elps.201700351.

[16]S.Ahrar,M.Hwang,P.N.Duncan,E.E.Hui,Analyst139(1)(2014)187,doi: http://dx.doi.org/10.1039/c3an01710a.

[17]X.Niu,F.Gielen,J.B.Edel,A.J.DeMello,Nat.Chem.3(6)(2011)437,doi:http:// dx.doi.org/10.1038/nchem.1046.

[18]M.R.Bringer,C.J.Gerdts,H.Song,J.D.Tice,R.F.Ismagilov,Philos.Trans.R.Soc.A Math.Phys.Eng.Sci.362(1818)(2004)1087,doi:http://dx.doi.org/10.1098/ rsta.2003.1364.

[19]I.Shestopalov,J.D.Tice,R.F.Ismagilov,LabChip4(4)(2004)316,doi:http://dx. doi.org/10.1039/b403378g.

[20]X.Niu,S.Gulati,J.B.Edel,A.J.Demello,LabChip8(11)(2008)1837,doi:http:// dx.doi.org/10.1039/b813325e.

[21]W.Postek,T.S.Kaminski,P.Garstecki,Analyst142(16)(2017)2901,doi:http:// dx.doi.org/10.1039/c7an00679a.

[22]S.Zeng,B.Li,X.Su,J.Qin,B.Lin,LabChip9(10)(2009)1340,doi:http://dx.doi. org/10.1039/b821803j.

[23]S.Jambovane,D.J.Kim,E.C.Duin,S.K.Kim,J.W.Hong,Anal.Chem.83(9)(2011) 3358,doi:http://dx.doi.org/10.1021/ac103217p.

[24]N.LiJeon,H.Baskaran,S.K.W.Dertinger,G.M.Whitesides,L.VanDeWater,M. Toner, Nat. Biotechnol. 20 (8) (2002) 826, doi:http://dx.doi.org/10.1038/ nbt712.

[25]W.Saadi,S.J.Wang,F.Lin,N.L.Jeon,Biomed.Microdevices8(2)(2006)109, doi:http://dx.doi.org/10.1007/s10544-006-7706-6.

[26]S.K.W.Dertinger,X.Jiang,Z.Li,V.N.Murthy,G.M.Whitesides,Proc.Natl.Acad. Sci. U. S. A. 99 (20) (2002) 12542, doi:http://dx.doi.org/10.1073/ pnas.192457199.

Fig.6. InfluenceoftheconcentrationandthetemperatureonthefluorescencesignalofRDdroplets.(A)FluorescenceintensityasafunctionofthetemperatureatvariousRD concentrations.(B)Rateoffluorescenceintensitydecrease(%)pertemperature(_C).

[27] C.Kim,K.Lee,J.H.Kim,K.S.Shin,K.J.Lee,T.S.Kim,J.Y.Kang,LabChip8(3) (2008)473,doi:http://dx.doi.org/10.1039/b714536e.

[28]T.M.Keenan,A.Folch,LabChip8(1)(2007)34,doi:http://dx.doi.org/10.1039/ b711887b.

[29]Y.Xia,G.M.Whitesides,Annu.Rev.Mater.Sci.28(1)(1998)153,doi:http://dx. doi.org/10.1146/annurev.matsci.28.1.153.

[30]Y.Yang,H.S.Rho,M.Stevens,A.G.J.Tibbe,H.Gardeniers,L.W.M.M.Terstappen, LabChip15(22)(2015)4331,doi:http://dx.doi.org/10.1039/c5lc00816f. [31] H.S.Rho,Y.Yang,A.T.Hanke,M.Ottens,L.W.M.M.Terstappen,H.Gardeniers,

LabChip16(2)(2016)305,doi:http://dx.doi.org/10.1039/c5lc01206f. [32]H.S.Rho,H.W.Veltkamp,A.T.Hanke,M.Ottens,C.Breukers,P.Habibovic,H.

Gardeniers, Molecules 25 (6) (2020) 1380, doi:http://dx.doi.org/10.3390/ molecules25061380.

[33]H.S.Rho,Y.Yang,H.-W.Veltkamp,H.Gardeniers.DirectDeliveryofReagents fromaPipetteTiptoaPDMSMicrofluidicDevice.RSCChips&Tips.Available at:https://blogs.rsc.org/chipsandtips/2015/10/09/

[34]R.Samy,T.Glawdel,C.L.Ren,Anal.Chem.80(2)(2008)369,doi:http://dx.doi. org/10.1021/ac071268c.

[35]D.Kim,H.S.Rho,S.Jambovane,S.Shin,J.W.Hong,Rev.Sci.Instrum.9(10) (2016)1340,doi:http://dx.doi.org/10.1063/1.4940927.

[36]T.Karstens,K.Kobs,J.Phys.Chem.84(14)(1980)1871,doi:http://dx.doi.org/ 10.1021/j100451a030.

[37]G.Arnaoutakis,D.Näther,QuenchingofFluorescencewithTemperature,(2016). [38]T.P.J.Knowles,D.A.White,A.R.Abate,J.J.Agresti,S.I.A.Cohen,R.A.Sperling,E.J. DeGenst,C.M.Dobson,D.A.Weitz,Proc.Natl.Acad.Sci.U.S.A.108(36)(2011) 14746,doi:http://dx.doi.org/10.1073/pnas.1105555108.