densitycpanel

ComparisonofjetsfromNewtonianandnon-Newtonianuidsinducedbyblister-actuatedlaser-inducedforwardtransfer(BA-LIFT)EmreTurkoz1LucDeike1,2CraigB.
Arnold1Received:27June2017/Accepted:8September2017Springer-VerlagGmbHGermany2017AbstractBlister-actuatedlaser-inducedforwardtransfer(BA-LIFT)isahigh-resolutionprintingtechnique,wheresmalldropletsareejectedfromathinliquidlayerontoareceiversubstrate.
Experimentswithahigh-speedcameraimagingsetupdemonstratenovelregimesduringthejetformationforthetransferofviscoelasticshear-thinningpolymersolutionscomparedtoNewtonianuids.
TheratiooftheinklmthicknessHftotheblisterheightHbisusedasadimensionlessnumberHf=Hbtoclassifydifferentjetbehaviors.
WeshowthatdifferentHf=Hbthresholdscanbedeneddependingontheelasticityoftheinklayerfortheinducedjetstoresultinbreakup.
1IntroductionLaser-inducedforwardtransfer(LIFT)isanozzle-lessprintingtechnique[1]thathasbeendevelopedovertheyearstoprintvarioustypesofmaterialsincludingpolymers[2],biomaterials[3],conductingsilverpastes[4],thinmetallms[5],carbonnanotubes[6],andviscoelasticalginatesolutions[7].
Blister-actuatedlaser-inducedfor-wardtransfer(BA-LIFT)isavariationofLIFTwherethereisaninterfacialpolymerlayerbetweenthetransparentsubstrateandtheinklayer[8].
Thistechniquehasbeendevelopedtopreventpossibledisruptionsinmechanicalandchemicalintegrityduetolaser–inkinteractions,andproventobesuccessfulintransferringsensitivematerials[9].
Thephysicsofblisterformationforthisprocesshasbeeninvestigatedexperimentally[10]andnumerically[11],andithasbeenshownthata355nmwavelengthlaserpulse(20ns)withaGaussianbeamprolefocusedonapolyimidethinlayerresultsinreproducibleblisterswhoseheightandradiusvaluesarefunctionsofthelaserenergy.
ThejetformationfromNewtonianuidshasbeeninves-tigatedexperimentallyusingtime-resolvedimaging[12]andnumericallyusingcomputationaluiddynamics(CFD)[13].
Inaddition,ananalyticalmodel[14]fortheearlytimedynamicsoftheBA-LIFTprocessshowsthattheuidaroundtheblisterundergoesshearastheinkatthevicinityoftheblisterispulledtowardsthecenteroftheblisterduringjetformation.
WhileBA-LIFTstudiessofarhavefocusedontheevaluationoftheunderlyingphysicswithjetscreatedusingNewtonianuids,real-lifeapplicationsrequiretheadap-tationofthistechniquetovarioustypesofmaterialsthatexhibitcomplexrheologicalproperties[15,16].
Inthisstudy,wepresentthefeaturesofjetsinducedusingvis-coelasticinkswiththeBA-LIFTtechnique.
ComparedtothejetscreatedfromNewtonianuids,thejetsfromvis-coelasticinksexhibitaveryhighstretchabilityduetotheirelasticity,whichcompeteswiththesurfacetension-driventhinning[17].
Wereportthatthereexistsacriticaljetvelocitythatwouldresultinthebreakupofjets.
2ExperimentalmethodsDifferentconcentrationsofxanthangum(XG)inwaterareusedasmodelnon-Newtoniansolutions.
XGsolutionisselectedastherheologicalmodiesbecauseoftheexten-siveliteraturededicatedtostudyingthesesolutions,its&CraigB.
Arnoldcbarnold@princeton.
edu1DepartmentofMechanicalandAerospaceEngineering,PrincetonUniversity,Princeton,NJ,USA2PrincetonEnvironmentalInstitute,PrincetonUniversity,Princeton,NJ,USA123Appl.
Phys.
A(2017)123:652DOI10.
1007/s00339-017-1252-3inherentviscoelasticity,andtheconvenientshear-thinningbehaviorofitsviscosity[18].
WeprepareXGinwatersolutionswithfourdifferentconcentrations([XG]=0.
05,0.
1,0.
2,and0.
4wt%).
TritonX-100surfactant(1.
0wt%)isaddedtothesolutiontoincreasethewettabilityofXGsolutiononthepolyimide.
Thesolutionsarestirredfor24hbeforetherheologicalmeasurementsandBA-LIFTexperimentstoensurehomogeneity.
ThemodelNewtonianuidusedinthisstudyisN-methyl-2-pyrrolidone(NMP)withtheviscosityl1:7mPas,thedensityq1030kg/m3,andthesurfacetensionc40:79mN/m.
TheshearrheometryofsolutionsisevaluatedusingAntonPaarPhysicaMCR-301rheometerusingadouble-gapgeometry.
Smallamplitudeoscillatoryshearrheometryinlinearviscoelasticregimeisutilizedtoevaluatetheyieldstressofthesolutions.
Theinterfacialpolyimidelayerispreparedbyspincoatingpolyimideresin(HDMicrosystemsPI2525)ontoglassmicroscopeslidesat500rpmfor10sfollowedbyafasterspinat3000rpmfor40s.
Subsequently,theslidesarebakedat120Cfor30minfollowedbyanother30minat360Ctocompletetheimidizationprocess,whichyields6.
9lm-thickpolyimidelayersonglassslides.
Theinksarecoatedonthepolymerlayerusingabladecoater.
Thethin-lmthicknessesaremeasuredusingaconfocalmicroscope.
Afrequency-tripledNd:YVO4laser(CoherentAVIA,20ns)isusedtogenerateapulsewith355-nmwavelength[10].
Thelaserbeamdiameterisapproximately20lm.
High-speedvideosarecapturedusingaPhantomv2512(VisionResearch)camerawitharangeof500,000–700,000framespersecondsand0.
26px/lmresolutionthankstoa10objective(Mitutoyo).
Thecameraexposuretimeis1.
0ls.
Sincethehigh-speedcameraworksinapost-triggercapturesettingwithacontinuousbacklight,therstimagewherethelaserpulseorajetfeatureisvisibleisdesignatedast0,whichdenotesthereferencetime.
Withtheuseofthehighspeedofthecamera,wecouldobtainmultipleimagesduringthejetmotion,andanalyzetheformationofjetsfromnon-Newtonianuids.
3Results3.
1CharacterizationofinksXGsolutionsexhibitshear-thinningandtheirshearviscosityvaluescanbeexpressedintheformofapowerlawl_ca_cn,where_cistheshearrate,nistheexponent,andaisthepowerlawcoefcient.
Theshear-thinningexponentdecreasesandthezero-shearviscosityincreaseswiththeXGconcentration([XG]).
Theyieldstresssyisevaluatedusingshearoscillatoryrheometrybydoingastresssweepat1Hzandreadingthecross-overofelastic(G0)andloss(G00)moduli.
ThesevaluesarepresentedinTable1alongwiththepowerlawparametersforXGsolutionsusedinthisstudy.
Theadditionofsurfactantaffectsthesurfacetensionofthesolutionsdrastically.
Usingapendantdropmethod,thesurfacetensionofthesolutionsisevaluatedas30.
6mN/m.
DuetothelowconcentrationofXG,thesolutionscanbeassumedtohavethesamedensityaswater(qw1000kg/m3).
NMPandXGsolutions,therefore,haveverysimilardensityandsurfacetensionvalues.
Furthermore,sinceXGsolutioniswater-basedandshear-thinning,itsviscosityapproacheswaterviscosity(lw1:002mPas)athighshear-ratesassociatedwiththeBA-LIFTprocess.
ThebiggestdifferencebetweenthesetwoinksistheinherentelasticityoftheXGsolutions.
3.
2BlisterprolesBlistersaregeneratedonthepolyimidelmusingalaserpulsethathasaGaussian-likeprole.
TheabsorptionprocessandtheresultingblisterformationaredepictedinFig.
1a.
Thelaserpulsecausesthedecompositionofthepolyimidelayer,whichleadstotheablationofpolyimidefromthetransparentglasssubstrate.
WhiletheinitialinklayerthicknessisdenotedwithHf,theradiusandtheheightoftheresultingblistersaredenotedwithRbandHb,respectively.
BlisterprolesmeasuredusingaconfocalmicroscopeareplottedinFig.
1b.
Figure1bshowsthattheblisterheightHbincreaseswiththelaserenergy.
Sincethepulsedurationisconstantfromonepulsetothenext,largerblisterheightisanindicationofthefasterblisterexpan-sion.
Keepingthebeamdiameterconstantandchangingthelaserenergybetween14.
4and28.
2lJ,theheightoftheblistercanbeassumedtobelinearlydependentonthelaserenergyasHb0:6684E0:78(inlm),whereEisthelaserenergy.
Ontheotherhand,theinkthicknessHfisdirectlyproportionaltotheresistanceduetoinertiatotheblisterexpansion.
ItwasshownthatthethresholdlaserenergyformaterialtransferdependsontheinkthicknesslinearlyforNewtonianinks[13].
Thus,wedeneadimensionlessparameterasHf=HbtocomparetheeffectofTable1Coefcientsofthepowerlawl_ca_cnofviscosityversusshearratewithyieldstresssyvalues[XG](%)ansy(Pa)0.
050.
269-0.
6070.
60.
10.
365-0.
6421.
40.
21.
122-0.
7402.
10.
42.
14-0.
7724.
0652Page2of6E.
Turkozetal.
123theuidinkthicknesstotheenergydepositedintothesystem.
WenotethattheHf=Hbparameterdoesnotincludeanyinformationabouttheradiusandtheshapeoftheblister,andamoregeneralanalysiswhichincludesdif-ferentblistershapesandradiiwouldrequiretheutilizationofadifferentdimensionlessparameter.
3.
3ImagesofBA-LIFTjets3.
3.
1Comparisonofjetimagesfromnon-NewtonianandNewtonianuidsFigure2a,bshowsjetsinducedfroma57lm-thickNMPink,whereahighenergyblister(Hb16:2lm,Hf=Hb3:51)resultsinmultipledropletformation(Fig.
2a),andalowenergyblister(Hb8:6lm,Hf=Hb6:63)resultsinjettingwithoutbreakup(Fig.
2b).
Weobservethatthejetisfasterandstretchesmoretobreakupintomultipledropletsthanthejettingwithoutbreakupcase.
ThejetpresentedinFig.
2areaches430.
4lmin1.
82ls,whilethejetinFig.
2breachesonly52lmin14.
54ls.
Multiple-dropbreakupandjettingwithoutbreakupcasesforXGsolutionsarepresentedinFig.
3a,b,respectively.
Theseguresshowthemultiple-dropbreakupandjettingwithoutbreakupregimesobtainedfor[XG]=0.
05wt%.
Bothmultiple-dropandjettingwithoutbreakupregimesarequalitativelyverydifferentfromtheseregimesobservedwithXGsolutions.
ItisseenfromFig.
3athatevenifthenon-NewtonianlamentstretchesuptoasimilaramountastheNewtonianlamentinFig.
2a,thenon-Newtonianl-amentdoesnotpinch-offfromthebaseofthejetastheNewtonianjetdoesatt027:27ls,butinstead,thenon-Newtonianjetbreaksacertaindistanceawayfromitsbasethatislargerthanitsradius.
Inaddition,wedonotseetheformationofmultipledropletsfromthelongnon-Newto-nianlamentasisthecasewiththeNewtonianlamentatFig.
1Absorptionofthelaserenergybythepolyimidelayerandtheblistergeneration.
aDimensionlessparameterHf=Hbisutilizedtodenotetherelativeeffectoftheblisterexpansiontotheresistanceagainstthejetformationbythebulkuid.
bBlisterprolescorrespondingtodifferentlaserenergiesmeasuredbylaserscanningmicroscopyFig.
2Twodifferentcasesofhigh-speedimagingwithNMP(Hf57lm):ahighenergyblister(Hb16:2lm,Hf=Hb3:51)resultinginbreakupwithmultipledrops,blowenergyblister(Hb8:6lm,Hf=Hb6:63)resultinginjettingwithoutbreakup.
Imagesarefromthevideoscapturedbythehigh-speedcamera,andt0isthereferencetime.
Scalebarsrepresent75lmComparisonofjetsfromNewtonianandnon-Newtonianuids…Page3of6652123t050:91ls.
Thiscomparisonshowsusthatthevis-coelasticityreducesthenumberofdropletsproducedfromasinglelament.
Thismatchestheobservationpresentedinaworkonthefragmentationofviscoelasticlaments[19],whereitisshownthatfewerdropletsareformedfromviscoelasticlamentscomparedtoNewtonianlaments.
TheinherentviscoelasticityofXGinksslowstheradialthinningofthejet,whichcausesfewerdropletscomparedtoajetformedfromNMPinks.
Wealsoobservethatthedropletisgettinglargerasthejetisretracted.
ComparingFig.
2bwithFig.
3bindicatesthatthefea-turesforthejettingwithoutbreakupareverydifferentbetweenthesetwotypesofuid.
Weobservethatthenon-Newtonianjetcreatedusing[XG]=0.
05wt%solutioncanbestretchedmorethantheNewtonianjetcreatedusingNMPanditstillcanretractwithoutproducingadroplet.
TheNMPjetinFig.
2bstretches91%oftheinklmthickness,whiletheXGjetstretches549%oftheinklmthickness.
3.
3.
2StretchingandelasticityofjetsfromxanthangumsolutionsAnimportantfeatureobservedduringtheexperimentsistheamountofstretchingassociatedwithXGconcentra-tions.
Weobservethatastheyieldstresssyoftheuidincreases,theamountofmaximumstretchingdecreasesforsimilarHf=Hbvalues.
ThisisshowninFig.
4a–c,whereimagesforthejettingwithoutbreakupregimefor[XG]=0.
1,0.
2,and0.
4wt%arepresented,respectively.
ItisshownintheseguresthatforsimilarHf=Hbvalues,theamountofstretchingdecreasesasthepolymerconcentra-tionincreases.
Inthesegures,thejetsstretchupto842,339,and243%oftheirinitialinkthicknessvaluesfor[XG]=0.
1,0.
2,0.
4wt%,respectively.
WeseeinFig.
4athatfor[XG]=0.
1wt%andHf=Hb5:65,adropletisformedatthetipofthejet.
Thisdropletgrowsasthejetgetsretractedtotheink,whichindicatesthatthereisapointalongthelamentwheretheaxialjetvelocityiszero.
Figure4bshowsimagesforthe[XG]=0.
2wt%andHf=Hb5:06case.
Asthejetretractsback,weseetheformationofashoulder-likestructurearound36.
9ls.
Thisstructurestaysasthejetisretracted,whicheventuallyroundsupandremainsvisibleaslongas167.
7ls.
Figure4cshowsimagesforthe[XG]=0.
4wt%andHf=Hb5:63.
Weseethattheleastamountofstretchingisobservedatthiscase.
Thisisexpected,because[XG]=0.
4wt%solutionshavethehighestyieldstress(sy4:0Pa)amongallthesolutionsusedinthisstudy.
Inthiscase,wedonotseetheformationofashoulderveryFig.
3Twodifferentcasesofhigh-speedimagingwith[XG]=0.
05wt%:ahighenergyblister(Hb16:2lm,Hf=Hb6:80)resultinginbreakupwithmultipledrops,blowenergyblister(Hb8:6lm,Hf=Hb8:65)resultinginjettingwithoutbreakup.
Imagesarefromthevideoscapturedbythehigh-speedcamera,andt0isthereferencetime.
Scalebarsrepresent75lmFig.
4Differentamountsofstretchingversuselasticity.
Higherelasticityaffectstheamountofstretching.
a[XG]=0.
1wt%,Hf=Hb5:65.
b[XG]=0.
2wt%,Hf=Hb5:06.
c[XG]=0.
4wt%,Hf=Hb5:63.
Imagesarefromthevideoscapturedbythehigh-speedcamera,andt0isthereferencetime.
Scalebarsrepresent75lm652Page4of6E.
Turkozetal.
123clearly;however,theretractedjetroundsupandisstillvisibleat149.
2ls.
AsampleofjetlengthversustimecurvesisplottedinFig.
5.
Exceptforthersttwoseries(Hf=Hb5:14andHf=Hb6:80with[XG]=0.
05wt%),allthedatapointsbelongtothecaseswherejetseventuallyretractbacktotheinklayersurface.
Wenotethatthetwocaseswithbreakuphaveadistinctivelyhighervelocitycomparedtothecaseswithoutbreakup.
Inaddition,weseethattheamountofstretchingincreasesastherelativeinkthicknessHf=Hbisdecreased.
Fromourexperiments,weseethatthethresholdHf=HbvaluesforbreakupcanbelistedasgiveninTable2.
ThisalsosuggeststhatacriticalvelocityorWebernumber(WeqU2jetRjet=r,whereqistheinkdensity,Ujetistheaxialjetvelocity,Rjetisthejetradius,andristhesurfacetension)mightbenecessarytoobtaindropsonviscoelasticligamentsasinthecaseofNewtonianlaments[20],whilethisthresholdmightdependontheyieldstressofthevis-coelasticlament.
Weobservefromourexperimentsthattherearetwomainfactorsthatdeterminethefeaturesoftheviscoelasticjets:(1)relativeinkthicknesscomparedtotheblisterheight(Hf=Hb);(2)polymerconcentration,whichaffectstheyieldstresssyandviscosityoftheink.
Theincreasinginkthicknessresultsinlargerinertialresistanceasexplainedin[13],wherethethresholdenergyforjetfor-mationhasbeenshowntobelinearlydependentontheinkthicknessforNewtonianinks.
Inadditiontotheinertialeffects,forourexperiments,theelasticityandhighvis-cosityoftheinklayercomeintothepicture.
Duetotheshear-thinningbehavioroftheinklayer,theeffectivevis-cosityoftheinkdecreasesastheuidfromthevicinityoftheblistergetspulledintowardsthecenteroftheblister[14].
Afterjetformation,theliquidlamentresiststhemotionintheaxialdirection.
Theresistanceisproportionaltotheelasticityofthelament;therefore,ahigherlaserenergyisrequiredtostretchlamentswithhigherelasticity.
4ConclusionsInsummary,thisworkpresentsnewinsightsintothemechanismsoflaser-inducedejectionsofnon-Newtonianuids.
WedemonstratedthatrelativetoNewtonianuidswithcomparablesurfacetension,density,andviscosity,thejetsinducedfromviscoelasticinkscanstretchsignicantlyfurtherbeforebreakup,wherebytheamountofstretchingdependsontheelasticityandtherelativethicknessHf=Hboftheinklayer.
Athresholdlaserenergyatwhichtheviscoelasticuidcanstretchtothepointofbreakupisfound,andthislaserenergydependsontheyieldstressoftheuid.
AlthoughtheexperimentalsetupusedinthisstudyisbasedonBA-LIFT,theresultsandinterpretationscanapplytootherLIFTtechniques.
AcknowledgementsWeacknowledgesupportoftheNationalSci-enceFoundation(NSF)throughaMaterialsResearchScienceandEngineeringCenterprogramthroughthePrincetonCenterforCom-plexMaterials(DMR-1420541).
References1.
C.
B.
Arnold,P.
Serra,A.
Pique,MRSBull.
32(01),23(2007)2.
A.
Palla-Papavlu,V.
Dinca,C.
Luculescu,J.
Shaw-Stewart,M.
Nagel,T.
Lippert,M.
Dinescu,J.
Opt.
12(12),124014(2010)3.
B.
Hopp,T.
Smausz,N.
Kresz,N.
Barna,Z.
Bor,L.
Kolozsvari,D.
B.
Chrisey,A.
Szabo,A.
Nogradi,TissueEng.
11(11–12),1817(2005)4.
C.
Boutopoulos,I.
Kalpyris,E.
Serpetzoglou,I.
Zergioti,Microuid.
Nanouid.
16(3),493(2014)5.
A.
I.
Kuznetsov,C.
Unger,J.
Koch,B.
N.
Chichkov,Appl.
Phys.
A106(3),479(2012)6.
A.
Palla-Papavlu,M.
Dinescu,A.
Wokaun,T.
Lippert,Appl.
Phys.
A117(1),371(2014)7.
Z.
Zhang,R.
Xiong,R.
Mei,Y.
Huang,D.
B.
Chrisey,Langmuir31(23),6447(2015)8.
N.
T.
Kattamis,P.
E.
Purnick,R.
Weiss,C.
B.
Arnold,Appl.
Phys.
Lett.
91(17),171120(2007)Fig.
5Jetlengthversustimeplotforvarioustestcases.
Thefastesttwocases(Hf=Hb5:14andHf=Hb6:80for[XG]=0.
05wt%)resultwithpinch-off,whileothersrepresentcaseswherejetsretractbacktotheinksurfaceTable2ThresholdHf=HbvaluesforbreakupevaluatedfordifferentXGconcentrations[XG](%)syHf=Hb0.
050.
2696.
830.
10.
3652.
480.
21.
1222.
180.
42.
140.
53ComparisonofjetsfromNewtonianandnon-Newtonianuids…Page5of66521239.
N.
T.
Kattamis,N.
D.
McDaniel,S.
Bernhard,C.
B.
Arnold,Organ.
Electron.
12(7),1152(2011)10.
M.
S.
Brown,N.
T.
Kattamis,C.
B.
Arnold,J.
Appl.
Phys.
107(8),083103(2010)11.
N.
T.
Kattamis,M.
S.
Brown,C.
B.
Arnold,J.
Mater.
Res.
26(18),2438(2011)12.
M.
S.
Brown,N.
T.
Kattamis,C.
B.
Arnold,Microuid.
Nanouid.
11(2),199(2011)13.
M.
S.
Brown,C.
F.
Brasz,Y.
Ventikos,C.
B.
Arnold,J.
FluidMech.
709,341(2012)14.
C.
F.
Brasz,C.
B.
Arnold,H.
A.
Stone,J.
R.
Lister,J.
FluidMech.
767,811(2015)15.
M.
Duocastella,J.
Fernandez-Pradas,P.
Serra,J.
Morenza,Appl.
Phys.
A93(2),453(2008)16.
V.
Dinca,A.
Patrascioiu,J.
Fernandez-Pradas,J.
Morenza,P.
Serra,Appl.
Surf.
Sci.
258(23),9379(2012)17.
A.
Ardekani,V.
Sharma,G.
McKinley,J.
FluidMech.
665,46(2010)18.
M.
Zirnsak,D.
Boger,V.
Tirtaatmadja,J.
Rheol.
43(3),627(1999)19.
B.
Keshavarz,E.
C.
Houze,J.
R.
Moore,M.
R.
Koerner,G.
H.
McKinley,Phys.
Rev.
Lett.
117(15),154502(2016)20.
J.
Hinze,AIChEJ.
1(3),289(1955)652Page6of6E.
Turkozetal.
123