synthesis of inorganic nanoparticles using microfluidic...
TRANSCRIPT
1
Synthesisofinorganicnanoparticlesusingmicrofluidicdevices
RazwanBaber
Athesissubmittedforthedegreeof
DoctorofPhilosophy
From
UniversityCollegeLondon
DepartmentofChemicalEngineering
Supervisors:
Prof.AsteriosGavriilidis
Prof.NguyenTKThanh
April2017
2
DeclarationofAuthorshipandOriginality
I,RazwanBaber,confirmthattheworkpresentedinthisthesisismyown.Whereinformationhasbeenderivedfromothersources,Iconfirmthatthishasbeenindicatedinthethesis.
Signature:……………………………………………………………………………..Date:April2017
3
Abstract
Thewell-knownbenefitsofusingmicro-scaleflowareassociatedwithcontrolledmasstransferbecause
of the reduced dimensions of microfluidic devices. The improvedmass transfer characteristics make
microfluidicdevicesanidealcandidateforsynthesisofnanoparticles(NPs).Inthisstudy,masstransfer
characteristicsare investigated formicrofluidicdevices suchas the coaxial flow reactor (CFR)and the
impingingjetreactor(IJR),throughtheuseoftheinteractionbyexchangewiththemeanmixingmodel,
Villermaux-Dushmanreactionscheme,highspeedcamera imagesandflowvisualizationusingdyeand
water. These reactors are contrasting in that the CFR has slowermixing because of its laminar flow
profilewhereasastheIJRhasmixingtimeintheorderofafewms.Theimportanceofmixingtimeas
wellasmanipulationofhydrodynamicsat themicroscale is shownthrough theuseof these typesof
reactorsforNPsynthesis.BothtypesofreactorshavebeneficialpropertiesforthesynthesisofNPs.NPs
were characterized using UV-Vis spectroscopy, DCS and TEM. Longer residence times and a higher
consumptionofreagents intheCFR ledtoreductionofsizeofsilverNPs, leadingtoreductions insize
from9.3±3nmto3.7±0.8nm.TheIJRshowedareductioninNPsizewithincreasingmixingefficiency,
withareductioninsizefrom7.9±5.8nmto3.4±1.4nmusingcitrateasaligandandfrom5.4±1.6nm
to4.2±1.1nmusingPVAasaligand.Asystemtocontrolnucleationandgrowthperiodswasdeveloped
bycombiningtheCFRwithacoiledflowinverter(CFI).SizecontrolovergoldNPswasachievedsimplyby
changingtheflowrateofreagents,withareductioninsizefrom23.9±4.7nmto17.9±2.1nm.Various
hydrodynamicswithin theCFRwerealso tested.Vortex flowathigherReynoldsnumbersdidnotgive
goodcontroloverNPsizeanddispersity,whilereducingtheinnertubeinternaldiameterresultedina
decreaseinNPsizefrom10.5±4.0nmto4.7±1.4nm.SilverandgoldNPsynthesiswasalsoperformed
inabatch reactor. Twodifferentmixing configurationsandvariation in theorderof reagentaddition
was investigated, and confirmed the importance ofmass transfer conditions in determining size and
dispersity.Inbothcases,fastandefficientmixingofthereducingagenttotheprecursorresultedinthe
smallestNPs.Therangeofsizesobtainedwere6.7±1.7nmto11.5±2.4nmand13.1±2.2nmto18.0±
4.8nmforsilverandgoldNPsrespectively.Alltheseresultsshowastronglinkbetweenmasstransfer
andNP sizeanddispersity for the systems tested, showing theneed forwell controlledandcarefully
consideredmasstransferconditions.
4
AcknowledgementsI would like to express my deepest gratitude to my supervisors, Prof. Asterios Gavriilidis and Prof.
Nguyen Thi Kim Thanh,whomwithout thisworkwould not have been possible. Prof. Gavriilidis, has
taughtmea lotabout research,buthasalsogivenme invaluableadviceoutsideofworkwhich really
helpedmotivatemewhentheresearchwasnotgoingmyway.Prof.Thanhwasinvaluableforherinput,
and along with her group at the Royal Institute of Great Britain, helped immeasurably with their
expertiseonnanoparticles.AspecialmentionalsogoestoDr.LucaMazzei,whohasbeenamentorto
meduringthisresearch.Hetaughtmewhatitmeanstobecriticalwhenassessinganyresearch,andto
neverstopbeingcurious.
Therearemanypeople to thank fromProf.Gavriilidis’ labgroup.Special thanks toEnhong,who
never ceases to amaze me with his flair for solving engineering problems. Thanks to Noor, Gaowei,
Redza,Achilleas,Luigi,Damiano,Hendrik,Anand,Huang,Spyros,Conor,CharaandValentina.Youhave
allhelpedmeinonewayoranother,whetheritwastakingthetimetochataboutaresearchproblemor
makingme smilewitha joke.Roxanne,Niall, Luke,AansandKuan,wewent throughmany trials and
tribulationstogetherwiththeevertemperamentalTEMthroughtheyearsandIthankyouforsharing.
Thanks tomygood friends,AnayetandAkhbar.Youare the funniestpeople Ihavemet,each in
yourownway.YoualwayslightenedmymoodwhenIwasbeingtooserious.
Myparents,MohammedandChandni,mysincerestappreciationgoesouttoyou.Youalwayskept
things in perspective forme, and reassuredmewhen Iwas unsure. You always gavememotivation,
whenIwasunmotivated,tokeepgoing.Thankyouforraisingme,IhopeImakeyouproudandcontinue
todosoeveryday.Tomybrothers,Arshad,AsadandImran,thankyouforteachingmethelessonsof
lifeandlookingoutformeeverystepoftheway.Tomysister,Yasmin,thankyouforteachingmehow
tobestronginadversity.Tomysister,Rehksena,thankyouforallthesupportandsacrificeyoumake
everysingledaytomakesureIamdoingwell.
Thankyoutomybest friend, foryour love,supportandunderstandingthroughoutthisPhD.You
helpedtokeepmegoingwhenIwasreadytogiveup.Youhaveperhapsmadethebiggestsacrificeofall
forputtingupwithme.Youarealwaysinmymindandheart.
Thankyouallformakingthisapleasantandenjoyablejourney.
5
ContentsAbstract........................................................................................................................................................3
Acknowledgements......................................................................................................................................4
Listoffigures................................................................................................................................................8
Listoftables...............................................................................................................................................16
Listofabbreviations...................................................................................................................................17
1 Introduction.......................................................................................................................................18
1.1 Motivation................................................................................................................................18
1.2 Nanoparticles............................................................................................................................18
1.3 Characterisationtechniquestodeterminethemorphologyofnanoparticles.........................19
1.3.1 Ultraviolet-visiblespectroscopy...........................................................................................19
1.3.2 Differentialcentrifugalsedimentation.................................................................................19
1.3.3 Transmissionelectronmicroscopy.......................................................................................19
1.4 Microfluidics.............................................................................................................................20
1.5 Mixing.......................................................................................................................................23
1.5.1 Diffusion...............................................................................................................................23
1.5.2 Advection.............................................................................................................................24
1.6 Objectives.................................................................................................................................24
1.7 Chapteroverview......................................................................................................................24
2 LiteratureReview..............................................................................................................................26
2.1 Microfluidicdevicesformasstransfer......................................................................................26
2.1.1 Parallellamination................................................................................................................26
2.1.2 Seriallamination...................................................................................................................26
2.1.3 Segmentedflow....................................................................................................................27
2.1.4 Injectionmicromixers...........................................................................................................27
2.1.5 Chaoticadvectionmicromixers............................................................................................27
2.2 Silvernanoparticles...................................................................................................................27
2.2.1 Batchsynthesisofsilvernanoparticlesviasodiumborohydridereduction.........................28
2.2.2 Microfluidicsynthesisofsilvernanoparticles......................................................................29
2.2.3 Mechanisticstudiesofsilvernanoparticlesynthesis...........................................................34
2.3 Goldnanoparticles....................................................................................................................39
2.3.1 Batchsynthesisofgoldnanoparticles..................................................................................39
2.3.2 Microfluidicsynthesisofgoldnanoparticles........................................................................43
2.3.3 Mechanisticstudiesofgoldnanoparticlesynthesis.............................................................48
3 Synthesisofsilvernanoparticlesinamicrofluidiccoaxialflowreactor............................................54
3.1 Experimental.............................................................................................................................55
3.1.1 Chemicals.............................................................................................................................55
3.1.2 ExperimentalSetup..............................................................................................................55
6
3.1.3 Synthesisofsilvernanoparticles..........................................................................................56
3.1.4 Characterizationofnanoparticles........................................................................................57
3.2 Investigationonvarianceinnanoparticlemorphologyusingparametricstudies....................57
3.2.1 Effectoftotalflowrate........................................................................................................57
3.2.2 Effectoftrisodiumcitrateconcentration.............................................................................60
3.2.3 Effectofsilvernitrateconcentration....................................................................................63
3.3 Conclusions...............................................................................................................................66
4 Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor.....................................68
4.1 Experimental.............................................................................................................................69
4.1.1 Chemicals.............................................................................................................................69
4.1.2 ExperimentalSetup..............................................................................................................70
4.1.3 NanoparticleSynthesis.........................................................................................................70
4.1.4 CharacterisationofNanoparticles........................................................................................70
4.2 Mixingcharacterizationmethods.............................................................................................71
4.2.1 Dilution-basedmethods.......................................................................................................71
4.2.2 Reaction-basedmethods......................................................................................................71
4.3 Villermaux-Dushmanreactionscheme.....................................................................................72
4.4 Theinteractionbyexchangewiththemeanmixingmodel......................................................72
4.5 Experimentaldeterminationofmixingtimeintheimpingingjetreactors..............................74
4.6 Flowregimecharacterizationofimpingingjetreactorusinghighspeedcamera....................77
4.7 InvestigatingtheeffectofwebernumberonNPmorphology.................................................79
4.7.1 EffectofwebernumberonsilverNPsynthesis1(citratesystem).......................................79
4.7.2 EffectofwebernumberonsilverNPsynthesis2(PVAsystem)...........................................85
4.8 Conclusions...............................................................................................................................91
5 Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice........................................................................................................93
5.1 Experimental.............................................................................................................................95
5.1.1 Chemicals.............................................................................................................................95
5.1.2 ExperimentalSetup..............................................................................................................95
5.1.3 Nanoparticlesynthesis.........................................................................................................97
5.1.4 Characterizationofnanoparticles........................................................................................98
5.2 Resultsanddiscussion..............................................................................................................98
5.2.1 Effectofincreasingflowrateongoldnanoparticlesizeanddispersity...............................98
5.2.2 Effectoftemperatureongoldnanoparticlesizeanddispersity........................................103
5.2.3 Silvernanoparticlesynthesisusingacoaxialflowdevicefollowedbyasplitandrecombinemixer 107
5.2.4 SilvernanoparticlesynthesisusingacoaxialflowdeviceoperatedathighRe..................110
5.2.5 EffectofinnertubeinternaldiameteronsilverNPsynthesisusingtheCFR.....................112
5.3 Conclusions.............................................................................................................................114
7
6 Masstransfereffectsinthebatchsynthesisofsilverandgoldnanoparticles................................116
6.1 Methodology..........................................................................................................................116
6.1.1 Chemicals...........................................................................................................................116
6.1.2 ExperimentalSetup............................................................................................................117
6.1.3 Nanoparticlesynthesis.......................................................................................................118
6.1.4 Characterizationofnanoparticles......................................................................................118
6.2 Resultsanddiscussion............................................................................................................119
6.2.1 MixingCharacterization.....................................................................................................119
6.2.2 Silvernanoparticlesynthesis..............................................................................................120
6.2.3 Goldnanoparticlesynthesis...............................................................................................125
6.3 Conclusions.............................................................................................................................130
7 Conclusionsandfurtherwork.........................................................................................................132
7.1 Conclusions.............................................................................................................................132
7.2 Furtherwork...........................................................................................................................133
AppendixA:MixingcharacterizationusingtheVillermaux-Dushmanreactions.....................................136
AppendixB:Synthesisofsilvernanoparticlesinamicrofluidiccoaxialflowreactor...............................142
AppendixC:Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor.......................144
AppendixD:Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice............................................................................................148
Listofpublications...................................................................................................................................152
Bibliography.............................................................................................................................................153
8
Listoffigures
Figure1-1:Schematicrepresentationofabatchreactor..........................................................................20
Figure1-2:Schematicrepresentationofaflowreactor............................................................................21
Figure1-3:Schematicrepresentationofturbulentflowandlaminarflow................................................21
Figure2-1:Exampleofaparallel laminationmixer.49Reproduced fromRef.49withpermissionofThe
RoyalSocietyofChemistry.........................................................................................................................26
Figure2-2:Slugswithamixtureofbluedyeandwatershowingadvectionwithin.ReprintedfromRef.
57,Copyright(2006),withpermissionfromElsevier.................................................................................27
Figure 2-3: Experimental flow reactor setup of silver NP synthesis used by Lin et al. Reprinted with
permissionfromRef.73.Copyright(2004)AmericanChemicalSociety....................................................29
Figure 2-4: Micro flow setup to synthesise silver nanorods and TEM image of silver nanorods.
ReproducedfromRef.74withpermissionofTheRoyalSocietyofChemistry..........................................30
Figure 2-5: Experimental setup of microreactor used by Liu et al. Reprinted from Ref. 79, Copyright
(2012),withpermissionfromElsevier.......................................................................................................31
Figure2-6:SpiralflowreactorusedbyKumaretal.,80Wrepresentsreactingphase,Krepresentsliquid
keroseneinertphase,Arepresentsgasairinertphase.Brepresentsinletmicromixer(crossmixer)used
togenerateslugs,withPEEKandStainlessSteelbeingused(dimensionsgiven).CandDarePMMA(0.5
mm ID) and Stainless Steel (1mm ID) spiral reactors respectively. Reprinted from Ref. 80, Copyright
(2012),withpermissionfromElsevier.......................................................................................................32
Figure 2-7: Configuration of microfludic device with four inlets: 1) Carrier fluid inlet, 2) Precursor
solutioninlet,3)Solvent inlet,4)Reducingagent inlet.Righthandsideshowssegmentedflowof ionic
liquid(BMIM-Tf2N)withintheoilcarrierphase(PCTFE).Athigherflowrates,jettingoccursfurtheralong
thechannelasthedropletspinchandformthinstrandphases.ReprintedwithpermissionfromRef.81.
Copyright(2012)AmericanChemicalSociety............................................................................................33
Figure2-8:SilverNPsmadeusingA)Microfluidicdevice,B)usingabatchreactor,C)UV-Viscomparison
ofboth (microfluidicdevice is black, batch is red). Reprintedwithpermission fromRef. 81.Copyright
(2012)AmericanChemicalSociety.............................................................................................................33
Figure2-9:Suggestedgrowthmechanism for silverNPsynthesis.Reprintedwithpermission fromRef.
68.Copyright(2012)AmericanChemicalSociety......................................................................................34
Figure 2-10:Mechanism of silver NP synthesis proposed by Takesue et al. 82 Silver nanoparticles are
formed through (I) an induction period during which Ag+ ions are reduced to Ag0 atoms; (II) a
nucleation-dominant formation period during which Ag13 clusters are nucleated and simultaneously
consumed in silver nanoparticle formation; and (III) a growth period during which larger silver
nanoparticlesareformedbycoalescenceandaggregationofsmallersilvernanoparticles.Reprintedwith
permissionfromRef.82.Copyright(2011)AmericanChemicalSociety....................................................35
9
Figure 2-11: Growth mechanisms proposed by Henglein and Giersig at low and optimal citrate
concentrationswithtwodifferentmechanismsoccurring.Atoptimalconcentrations,itissuggestedthat
asurfacereductionofsilverionsoccursonthesurfaceofformedNPsowingtothereductionpotential
of the citrate adsorbed onto the surface. At low concentrations, coalescence of smaller silver NP
clustersoccurs.34,83Ref.34publishedbyTheAmericanChemicalSociety...............................................36
Figure 2-12: Absorption spectra of Ag colloids prepared by γ irradiation of an aqueous solution (N2
purged)containing1mMAgNO3,1%methanol,and(a)1mM,(b)5mM,and(c)10mMsodiumcitrateas
stabilizer.84ReprintedwithpermissionfromRef.84.Copyright(2003)AmericanChemicalSociety........37
Figure 2-13: Growthmechanism through pulse radiolysis of amixture of silver nitrate and trisodium
citrate proposed by Pillai and Kamat: primary and secondary growth steps in the formation of silver
nanocrystallites.84ReprintedwithpermissionfromRef.84.Copyright(2003)AmericanChemicalSociety.
....................................................................................................................................................................37
Figure2-14:MechanismsofsilverNPgrowthusingdifferentbeamcurrentsunderaTEMmicroscope.86
ReprintedwithpermissionfromRef.86.Copyright(2012)AmericanChemicalSociety...........................38
Figure2-15:Microfluidicdeviceused tosynthesizehexagonalgoldNPs.40Reprinted fromRef.40with
permissionfromTheInstituteofPhysics...................................................................................................43
Figure2-16:SynthesisofgoldNPsusingsyringepumps,t-mixerandfusedsilicacapillarytubingsetina
hotplate.Reproduced(Adapted)fromRef.107withpermissionofTheRoyalSocietyofChemistry......44
Figure2-17:SplitandrecombinemixeremployedforthesynthesisofgoldNPs.ReproducedfromRef.
109withpermissionofTheRoyalSocietyofChemistry............................................................................45
Figure2-18:InterdigitatingmicromixerusedbyShalometal.forthesynthesisofgoldNPsviatheBrust-
Schiffrinmethod.110ReprintedfromRef.110,Copyright(2007),withpermissionfromElsevier.............46
Figure 2-19:Segmented flow in microchannel employed for synthesis of gold NPs by Cabeza et al.
ReprintedwithpermissionfromRef.111.Copyright(2012)AmericanChemicalSociety.........................47
Figure 2-20: Interdigitatingmixer used by Tsunoyama et al. for the synthesis of gold NPs. Reprinted
(adapted)withpermissionfromRef.112.Copyright(2008)AmericanChemicalSociety.........................47
Figure2-21:MechanismsuggestedbyJietal.forgoldNPsynthesis.Twopathwaysexistdependenton
pH.95ReprintedwithpermissionfromRef.84.Copyright(2007)AmericanChemicalSociety..................51
Figure 2-22: Mechanism of gold NP formation suggested by Polte et al. Reprinted (adapted) with
permissionfromRef.120.Copyright(2010)AmericanChemicalSociety..................................................51
Figure2-23:MechanismofgoldNPformationintheTurkevichsynthesissuggestedbyPolte.Ref.124–
PublishedbyTheRoyalSocietyofChemistry.............................................................................................52
Figure3-1:Schematicofthecoaxialflowreactorsetup. Insertshowsflowvisualizationof laminarflow
insidethecoaxialflowreactorwithbluedyeflowingthroughtheinnertubeandwaterflowingthrough
theoutertube............................................................................................................................................56
10
Figure 3-2: TEM images and particle size distributions of silver NPs synthesized at different total
flowrates.A:1ml/min,B:2.5ml/min,C:8ml/minandD:14ml/min.Concentrationofsilvernitrate0.2
mM,trisodiumcitrate0.2mM,sodiumborohydride0.6mM.0.798mminnertubeI.D..........................58
Figure 3-3: TEM images andparticle size distributionsof silverNPs synthesized at different trisodium
citrateconcentrations.A:0.05mM,B:0.1mM,C:0.25mMandD:1.5mM.Concentrationofsilver
nitrate0.1mM,sodiumborohydride0.3mM.Totalflowrate2.5ml/min,0.556mminnertubeI.D......62
Figure3-4:AbsorbancepeaksofsilverNPssynthesizedatvarioustrisodiumcitrateconcentrationsinthe
range 0.025- 1.5 mM. Total flow rate 2.5 ml/min, concentration of silver nitrate 0.1 mM, sodium
borohydride0.3mM.0.556mminnertubeI.D.........................................................................................63
Figure3-5:TEMimagesandparticlesizedistributionsofsilverNPssynthesizedatdifferentsilvernitrate
concentrations.A:0.05mM,B:0.15mM,C:0.25mMandD:0.4mM.Concentrationoftrisodiumcitrate
0.5mM,sodiumborohydride0.3mM.Totalflowrate2.5ml/min,0.556mminnertubeI.D..................65
Figure 3-6: Absorbance peaks of silver NP synthesized at various silver nitrate concentrations in the
range 0.05-0.4mM. Total flow rate 2.5ml/min, concentrations of trisodium citrate 0.5mM, sodium
borohydride0.3mM.0.556mminnertubeI.D. Inset:Silvernitrateconcentrationvs.peakabsorbance
forresonancepeaks...................................................................................................................................66
Figure4-1:Representationoftheimpingingjetreactor............................................................................70
Figure 4-2: Segregation index vs. mixing time estimated using the IEM model for the Villermaux-
Dushmannreactionschemefor0.5mm(blueline)and0.25mm(redline)IJRs.Concentrationsofbuffer
solutionwereKI:0.032M,KIO3:0.006M,NaOH:0.045M,H3BO3:0.09MinbothIJRs.Concentrationof
H2SO4was:0.017Mand0.02Mfor0.5mmand0.25mmIJRrespectively..............................................74
Figure4-3:MixingtimecalculatedbytheVillermaux-Dushmannreactionschemevs.Webernumberfor
0.5mm(blacksquares)and0.25mm(red triangles) IJRs.Concentrationofacidsolution (input1)was
0.017Mand0.02MH2SO4for0.5mmand0.25mmI.D.tubingIJRrespectively.Concentrationsofbuffer
solution(input2)wereKI:0.032M,KIO3:0.006M,NaOH:0.045M,H3BO3:0.09MinbothIJRs............75
Figure4-4:Flowvisualizationof jetsemittedfrom0.5mmtubesattotal flowrate:A:32ml/min(We:
13),B:40ml/min(We:20),C:48ml/min(We:29),D:56ml/min(We:39),E:64ml/min(We:51)andF:
72ml/min(We:65).Picturesshowsideviewandfrontview....................................................................78
Figure4-5:Flowvisualizationofjetsemittedfrom0.25mmtubesattotalflowrate:A:18ml/min(We:
32),B:22ml/min(We:48),C:26ml/min(We:68),D:30ml/min(We:90),E:34ml/min(We:115),F:38
ml/min(We:144)andG:42ml/min(We:176).Picturesshowsideviewandfrontview.........................79
Figure4-6: TEM imagesandparticle sizedistributionsof silverNPs synthesisedatdifferent total flow
ratesusinga0.5mmI.D.tubingIJR.(A)32ml/min(We:13),(B)48ml/min(We:29),(C)64ml/min(We:
51)and(D)72ml/min(We:65).Concentrationofsilvernitrate0.9mM,trisodiumcitrate6mMininput
1andsodiumborohydride1.8mMininput2...........................................................................................82
11
Figure4-7: TEM imagesandparticle sizedistributionsof silverNPs synthesisedatdifferent total flow
ratesusinga0.25mm I.D. tubing IJR. (A)20ml/min (We:40), (B)24ml/min (We:58), (C)28ml/min
(We:78)and(D)32ml/min(We:102).Concentrationofsilvernitrate0.9mM,trisodiumcitrate6mMin
input1andsodiumborohydride1.8mMininput2..................................................................................83
Figure4-8:Dependenceofpeakabsorbance(peakwavelength386-389nm)onWebernumberobtained
fromUV-VisspectroscopyofsilverNPssynthesisedatdifferenttotalflowratesusing: i)0.5mm(black
squares)andii)0.25mmI.D.tubingIJR(redtriangles).Concentrationofsilvernitrate0.9mM,trisodium
citrate6mMininput1andsodiumborohydride1.8mMininput2........................................................84
Figure4-9: TEM imagesandparticle sizedistributionsof silverNPs synthesizedatdifferent total flow
ratesusinga0.5mmI.D.tubingIJR.(A)32ml/min(We:13),(B)48ml/min(We:29),(C)64ml/min(We:
51)and (D)72ml/min (We:65).Concentrationofsilvernitrate0.9mM,PVA0.02wt% in input1and
sodiumborohydride1.8mMininput2.....................................................................................................88
Figure4-10:TEMimagesandparticlesizedistributionsofsilverNPssynthesizedatdifferenttotalflow
ratesusinga0.25mm I.D. tubing IJR. (A)18ml/min (We:32), (B)26ml/min (We:68), (C)34ml/min
(We:115)and(D)42ml/min(We:176).Concentrationofsilvernitrate0.9mM,PVA0.02wt%ininput1
andsodiumborohydride1.8mMininput2..............................................................................................89
Figure 4-11: Dependence of peak absorbance (peak wavelength 398-401 nm) on Weber number
obtained fromUV-Vis spectroscopyof silverNPs synthesisedat different total flow ratesusing: i) 0.5
mm (black squares) and ii) 0.25mm (red triangles) I.D. tubing IJR. Concentration of silver nitrate 0.9
mM,PVA0.02wt%ininput1andsodiumborohydride1.8mMininput2..............................................90
Figure4-12:DependenceofsilvernanoparticlediameterontheWebernumber.Datashownis for0.5
mmand0.25mmIJRsandforbothcitrateandPVAsurfactants...............................................................92
Figure5-1:Schematicrepresentationsofcomponentsusedinthemicrofluidicflowsetup:a:coaxialflow
reactor,b:coiledflowinverterandc:splitandrecombinemixer.............................................................94
Figure5-2:SchematicrepresentationofexperimentalsetupforgoldNPsynthesis.................................96
Figure 5-3: Schematic representation experimental setup for silver NP synthesis: a) CFR used in
conjunctionwithSARandb)CFRusedinisolation....................................................................................97
Figure5-4:TEMimagesofgoldNPssynthesizedusingtheCFRwithaCFIresidenceloopataflowrateof
A:0.25ml/min,B:0.5ml/min,C:0.75ml/min,D:1ml/min,E:1.5ml/min,F:2ml/minandG:3ml/min.
H:AveragediameterofgoldNPsvsflowrateofsynthesis(barsrepresentstandarddeviationofthesize).
Concentration of tetrachloroauric acid, 0.557 mM; concentration of trisodium citrate, 0.09 M;
volumetric flow rate ratio, 32.3: 1 (Qout:Qin, HAuCl4: Na3citrate); molar flow rate ratio, 1:5 (HAuCl4:
Na3citrate);temperature,80°C................................................................................................................101
Figure5-5:NormalizedDCScurvesofgoldNPssynthesizedusingtheCFRwithaCFI residence loopat
flow rates between 0.25 and 3ml/min. Inset: Diameter of gold NPs vs. flow rate of synthesis (bars
represent standard deviation of size). Concentration of tetrachloroauric acid was 0.557 mM and
12
concentration of trisodium citrate was 0.09 M. Concentration of tetrachloroauric acid, 0.557 mM;
concentration of trisodium citrate, 0.09 M; volumetric flow rate ratio, 32.3: 1 (Qout:Qin, HAuCl4:
Na3citrate);molarflowrateratio,1:5(HAuCl4:Na3citrate);temperature,80°C.....................................102
Figure5-6:UV-VisspectraofgoldNPssynthesizedusingtheCFRwithaCFIresidenceloopatflowrates
between 0.25 and 3 ml/min. Inset: Peak absorbance (black squares) and peak wavelength (red
diamonds)ofgoldNPsvs.flowrateofsynthesisanalyzedusingUV-Visspectroscopy.Concentrationof
tetrachloroauricacid,0.557mM;concentrationoftrisodiumcitrate,0.09M;volumetricflowrateratio,
32.3:1(Qout:Qin,HAuCl4:Na3citrate);molarflowrateratio,1:5(HAuCl4:Na3citrate);temperature,80°C.
..................................................................................................................................................................102
Figure 5-7: TEM images of gold NPs synthesized using the CFR with a CFI residence loop at a
temperature of A: 60°C, B: 70°C, C: 80°C, D: 90°C, E: 100°C. F: Average diameter of gold NPs vs
temperature of synthesis (bars represent standard deviation of the size). Concentration of
tetrachloroauricacid,0.557mM;concentrationoftrisodiumcitrate,0.09M;volumetricflowrateratio,
32.3:1(Qout:Qin,HAuCl4:Na3citrate);molarflowrateratio,1:5(HAuCl4:Na3citrate);flowrate,1ml/min.
..................................................................................................................................................................105
Figure5-8:NormalizedDCScurvesofgoldNPssynthesizedusingtheCFRwithaCFI residence loopat
temperatures between 60 and100°C. Inset: Diameter of gold NPs vs. temperature of synthesis (bars
representstandarddeviationofsize).Concentrationoftetrachloroauricacid,0.557mM;concentration
oftrisodiumcitrate,0.09M;volumetricflowrateratio,32.3:1(Qout:Qin,HAuCl4:Na3citrate);molarflow
rateratio,1:5(HAuCl4:Na3citrate);flowrate,1ml/min..........................................................................106
Figure 5-9: UV-Vis spectra of gold NPs synthesized using the CFR with a CFI residence loop at
temperaturesbetween60and100°C.Inset:Peakabsorbance(blacksquares)andpeakwavelength(red
diamonds)ofgoldNPsvs.temperatureofsynthesis.Concentrationoftetrachloroauricacid,0.557mM;
concentration of trisodium citrate, 0.09 M; volumetric flow rate ratio, 32.3: 1 (Qout:Qin, HAuCl4:
Na3citrate);molarflowrateratio,1:5(HAuCl4:Na3citrate);flowrate,1ml/min....................................106
Figure 5-10: TEM images of silverNPs synthesized using the CFR followed by a SARmixer at a silver
nitrate concentration of A: 0.05mM, B: 0.15mM, C: 0.25mM,D: 0.4mM. Concentration of sodium
borohydride,0.3mM;concentrationoftrisodiumcitrate,0.5M;volumetricflowrateratio,1:1(Qout:Qin,
NaBH4: AgNO3); molar flow rate ratio, 0.5-4:3:5 (AgNO3:NaBH4:Na3citrate); flow rate, 2.5 ml/min;
temperature,22-24°C..............................................................................................................................108
Figure5-11:UV-VisspectraofsilverNPssynthesizedusingtheCFRfollowedbyaSARmixeratasilver
nitrateconcentrations rangingbetween0.05and0.4mM. Inset:Peakabsorbance (blacksquares)and
peakwavelength(reddiamonds)ofsilverNPsvs.silvernitrateconcentrationofsynthesis.Concentration
ofsodiumborohydride,0.3mM;concentrationoftrisodiumcitrate,0.5M;volumetricflowrateratio,1:
1 (Qout:Qin, NaBH4: AgNO3); molar flow rate ratio, 0.5-4:3:5 (AgNO3:NaBH4:Na3citrate); flow rate, 2.5
ml/min;temperature,22-24°C.................................................................................................................109
13
Figure5-12:Diametervs.silvernitrateconcentrationfortheCFRoperatedfollowedwithaSAR(black
squares) and without a SAR (red diamonds). Concentration of sodium borohydride, 0.3 mM;
concentrationoftrisodiumcitrate,0.5M;volumetricflowrateratio,1:1(Qout:Qin,NaBH4:AgNO3);molar
flowrateratio,0.5-4:3:5(AgNO3:NaBH4:Na3citrate);flowrate,2.5ml/min;temperature,22-24°C......109
Figure 5-13: TEM image of silver NPs synthesized using the CFR with inner tube I.D. of 0.798 mm
operatedunderhighRewithatotalflowrateofA:5.1ml/min,B:10.1ml/min,C:15.1ml/min,D:20.1
ml/min.Theoutertubeflowratewasfixedat0.1ml/min.Concentrationofsilvernitrate,0.1mM;of
sodiumborohydride,0.3mM;concentrationof trisodiumcitrate,0.5M (all concentrations statedare
thoseaftercompletemixing);molar flowrate ratio,1:3:5 (AgNO3:NaBH4:Na3citrate); temperature,22-
24°C..........................................................................................................................................................111
Figure5-14:FlowvisualizationofCFRwith innertubeI.D.of0.798mmoperatedunderturbulentflow
conditions.BasicBluedyewaspumped through the inner tubeat20ml/minandwaterwaspumped
throughtheoutertubeat0.1ml/min.Reis520intheinnertubeand212inthemainchannel.Picture1
is takenbeforedye ispumped through the inner tubeup,Pictures2-5asoperational time increased,
Picture6aftersteadystateisreached.....................................................................................................112
Figure5-15:TEMimagesofsilverNPssynthesizedusingtheCFRwithinnertubediametersofA:0.142
mm,B:0.345mm,C:0.447mm,D:0.556mm,E:0.701mm,F:0.798mm,G:Nanoparticlediametervs.
inner tube internal diameter. Concentration of silver nitrate, 0.1 mM; concentration of sodium
borohydride,0.3mM;concentrationoftrisodiumcitrate,0.5M;volumetricflowrateratio,1:1(Qout:Qin,
NaBH4:AgNO3);molarflowrateratio,1:3:5(AgNO3:NaBH4:Na3citrate);flowrate,1ml/min,temperature
22-24°C....................................................................................................................................................113
Figure5-16:UV-Vis spectraof silverNPssynthesizedusing theCFRwith inner tubediameters ranging
between 0.142 and 0.798 mm. Inset: Peak absorbance (black squares) and peak wavelength (red
diamonds)ofsilverNPsvs.innertubeinternaldiameterusedintheCFR.Concentrationofsilvernitrate,
0.1 mM; concentration of sodium borohydride, 0.3 mM; concentration of trisodium citrate, 0.5 M;
volumetric flow rate ratio, 1: 1 (Qout:Qin, NaBH4: AgNO3); molar flow rate ratio, 1:3:5
(AgNO3:NaBH4:Na3citrate);flowrate,1ml/min,temperature22-24°C..................................................114
Figure6-1:SchematicofexperimentalsetupforsilverandgoldNPsynthesisconsistingofanErlenmeyer
flaskwith precursor or reducing agent stirred via amagnetic stirrerwith a syringe pump reagent via
tubinga)silverNPsystemusingastainlesssteel0.5mmI.D.tubingandb)goldNPsystemusingaPTFE
0.5mmI.D.tubingwithaglycerinebathandtemperatureprobefortemperaturecontrolat80°C.Stirrer
speedwassetat500rpminbothcases..................................................................................................117
Figure6-2:Schematicrepresentationofinjectionpointswithinthebatchvesseltovarymixingefficiency
a) injectionisaddedfromabovethesolutioninthebatchvesselandb) injectionisaddedclosetothe
stirbartip.................................................................................................................................................119
14
Figure 6-3: Segregation index vs. mixing time estimated using the IEM model for the Villermaux-
Dushmann reaction scheme. Concentrations of buffer solution were KI: 0.0117 M, KIO3: 0.0023 M,
NaOH:0.0909M,H3BO3:0.1818M.ConcentrationofH2SO4was:0.07M...............................................120
Figure6-4:TEMimagesofsilverNPssynthesized inabatchvesselundervariousconditions.A)0.6ml
sodiumborohydride[63.3mM]addedviadropsto20mlofamixtureofsilvernitrate [0.38mM]and
trisodium citrate [0.383mM] at a flow rate of 0.5ml/min, B) 0.6ml sodium borohydride [63.3mM]
addednearthestirbartipto20mlofamixtureofsilvernitrate[0.38mM]andtrisodiumcitrate[0.383
mM]ataflowrateof50ml/min,C)0.6mlofamixtureofsilvernitrate[9.1mM]andtrisodiumcitrate
[9.1mM]addedviadropsto20mlsodiumborohydride[1.36mM]ataflowrateof0.5ml/minandD)
0.6mlofamixtureofsilvernitrate[9.1mM]andtrisodiumcitrate[9.1mM]addednearthestirbartip
to20mlsodiumborohydride[1.36mM]ataflowrateof50ml/min.Experimentswereconductedat
roomtemperatureandstirringspeedwassetat500rpm......................................................................124
Figure6-5:AveragepeakabsorbanceandaveragepeakwavelengthobtainedusingUV-visspectroscopy
ofsilverNPssynthesizedinbatchvesselundervariousconditions.A)0.6mlsodiumborohydride[63.3
mM]addedviadropsto20mlofamixtureofsilvernitrate[0.38mM]andtrisodiumcitrate[0.383mM]
ataflowrateof0.5ml/min,B)0.6mlsodiumborohydride[63.3mM]addednearthestirbartipto20
mlofamixtureofsilvernitrate[0.38mM]andtrisodiumcitrate[0.383mM]ataflowrateof50ml/min,
C)0.6mlofamixtureofsilvernitrate[9.1mM]andtrisodiumcitrate[22.7mM]addedviadropsto20
ml sodiumborohydride [1.36mM] at a flow rate of 0.5ml/min andD) 0.6ml of amixture of silver
nitrate [9.1 mM] and trisodium citrate [22.7 mM] added near the stir bar tip to 20 ml sodium
borohydride[1.36mM]ataflowrateof0.5ml/min.Experimentswereconductedatroomtemperature
andstirringspeedwassetat500rpm.Barsrepresentstandarddeviationofabsorbanceandwavelength
for3repeatedexperiments.....................................................................................................................125
Figure6-6: TEM imagesof goldNPs synthesized inabatchvesselunder various conditions.A)0.6ml
trisodiumcitrate[92.7mM]addedviadropsto20mloftetrachloroauricacid[0.556mM]ataflowrate
of 0.5 ml/min, B) 0.6 ml trisodium citrate [92.7 mM] added near the stir bar tip to 20 ml of
tetrachloroauricacid [0.556mM]ata flow rateof50ml/min,C)0.6mlof tetrachloroauricacid [18.5
mM]addedviadropsto20mltrisodiumcitrate[2.78mM]ataflowrateof0.5ml/minandD)0.6mlof
tetrachloroauricacid[18.5mM]addednearthestirbartipto20mltrisodiumcitrate[2.78mM] ata
flowrateof50ml/min.Experimentswereconductedat80°Candstirringspeedwassetat500rpm..128
Figure 6-7: Peak absorbance and peak wavelength obtained using UV-vis spectroscopy of gold NPs
synthesizedinbatchvesselundervariousconditions.A-Au)0.6mltrisodiumcitrate[92.7mM]added
via drops to 20ml of tetrachloroauric acid [0.556mM] at a flow rate of 0.5ml/min, B - Au) 0.6ml
trisodiumcitrate[92.7mM]addednearthestirbartipto20mloftetrachloroauricacid[0.556mM]ata
flow rate of 50ml/min, C - Au) 0.6ml of tetrachloroauric acid [18.5mM] added via drops to 20ml
trisodiumcitrate[2.78mM]ataflowrateof0.5ml/minandD-Au)0.6mloftetrachloroauricacid[18.5
mM] added near the stir bar tip to 20ml trisodium citrate [2.78mM] at a flow rate of 0.5ml/min.
15
Experimentswere conductedat 80°Cand stirring speedwas set at 500 rpm.Bars represent standard
deviationofabsorbanceandwavelengthfor3repeatedexperiments...................................................129
Figure6-8:NormalizednumberpercentagevsdiameterobtainedusingDCSofgoldNPssynthesizedin
batchvesselundervariousconditions.A-Au)0.6mltrisodiumcitrate[92.7mM]addedviadropsto20
mlof tetrachloroauricacid [0.556mM]ata flow rateof0.5ml/min,B -Au)0.6ml trisodiumcitrate
[92.7mM]addednearthestirbartipto20mloftetrachloroauricacid[0.556mM]ataflowrateof50
ml/min,C -Au)0.6mlof tetrachloroauricacid [18.5mM]addedviadrops to20ml trisodiumcitrate
[2.78mM] ataflowrateof0.5ml/minandD -Au)0.6mloftetrachloroauricacid[18.5mM]added
nearthestirbartipto20mltrisodiumcitrate[2.78mM]ataflowrateof0.5ml/min.Experimentswere
conductedat80°Candstirringspeedwassetat500rpm.Insetisanaveragediameterforeachcondition
obtainedfrom3repeatexperiments.......................................................................................................130
16
Listoftables
Table3-1:FWHMandpeakwavelengthofsilverNPsatvarioustrisodiumcitrateconcentrations inthe
range 0.025- 1.5 mM. Total flow rate 2.5 ml/min, concentration of silver nitrate 0.1 mM, sodium
borohydride0.3mM.0.556mminnertubeI.D.........................................................................................63
Table3-2:FWHMandpeakwavelengthofsilverNPsatvarioussilvernitrateconcentrationsintherange
0.05 – 0.4 mM. Total flow rate 2.5 ml/min, concentrations of trisodium citrate 0.5 mM, sodium
borohydride0.3mM.0.556mminnertubeI.D.........................................................................................66
Table 4-1: FWHM and peak wavelength of silver NPs synthesised in the 0.5 mm IJR using trisodium
citrateasastabilisingagentatflowratesbetween32and72ml/min.Concentrationofsilvernitrate0.9
mM,trisodiumcitrate6mMininput1andsodiumborohydride1.8mMininput2................................84
Table 4-2: FWHMand peakwavelength of silverNPs synthesised in the 0.25mm IJR using trisodium
citrateasastabilisingagentatflowratesbetween18and34ml/min.Concentrationofsilvernitrate0.9
mM,trisodiumcitrate6mMininput1andsodiumborohydride1.8mMininput2................................85
Table 4-3: FWHM and peakwavelength of silver NPs synthesised in the 0.5mm IJR using PVA as a
stabilisingagentatflowratesbetween32and72ml/min.Concentrationofsilvernitrate0.9mM,PVA
0.02wt%ininput1andsodiumborohydride1.8mMininput2..............................................................90
Table 4-4: FWHMandpeakwavelengthof silverNPs synthesised in the 0.25mm IJRusing PVAas a
stabilisingagentatflowratesbetween18and34ml/min.Concentrationofsilvernitrate0.9mM,PVA
0.02wt%ininput1andsodiumborohydride1.8mMininput2..............................................................91
Table 5-1: Diameter and dispersity of gold NPs synthesised at various flow rates with corresponding
Pecletnumber..........................................................................................................................................103
Table6-1:SummaryofconcentrationsforeachmixingconditionusedinthesilverNPsynthesis.........121
Table6-2:SummaryofconcentrationsforeachmixingconditionusedinthegoldNPsynthesis...........126
17
Listofabbreviations
CFD
Computationalfluiddynamics
CFI
Coiledflowinverter
CFR
Coaxialflowreactor
DCS
DifferentialCentrifugalSedimentation
ETFE
Ethylenetetrafluoroethylene
FWHM
Fullwidthathalfmaximum
IJR
Impingingjetreactor
NP
Nanoparticle
PTFE
Polytetrafluoroethylene
PVA
Polyvinylalcohol
RTD
Residencetimedistribution
SAR
Splitandrecombine
SAXS
Smallanglex-rayscattering
TEM
Transmissionelectronmicroscopy
UV-Vis
Ultraviolet-visible
18
1 Introduction
1.1 Motivation
In recent years, nanoparticles (NPs) have become a widely researched area to investigate themany
interestingpropertiesarisingfromallmannerofdifferentmaterials,shapesandsizes.Whenconsidering
the“bottom-up”chemicalsynthesisapproachtomakingNPs,moststudiesemploytheuseofreactors
consistingofavesselwhichismechanicallystirred. Inchemicalengineeringterms,thesereactionsare
carried out in a “batch” reactor. The benefits of using such reactors suit the aim of discovering new
methods involvingdifferentconcentrations, temperaturesand reagents ina relatively simplemanner.
More recently, microfluidic devices have been employed to synthesize NPs because of the various
benefitsofferedbythesetypesofreactors.Theseincludeimprovedheatandmasstransfer,efficiency
and safety.1 It is believed because of these enhanced characteristics (in particular heat and mass
transfer); improvements in the morphological characteristics of NPs synthesized using chemical
methodscanbemade(i.e.narrowsizedistributionsandcontrolofsize).
1.2 Nanoparticles
NPsarematerials,withasizerangefrom1to100nm,whichdifferfromtheirbulkcounterpartsbecause
they exhibit uniqueproperties due to their high surface area to volume ratios andquantumeffects.2
Importantly, thepropertiesof theNPsdependonthesize,compositionandshape,makingthemvery
different from their bulk counterpartswhich retain their properties irrespectiveof size.NPs exhibit a
largesurfaceareatovolumeratio(alargepercentageoftheiratomsarelocatedonthesurface),which
changes significantly with increasing size in the scale between 1-100 nm, resulting in changing
propertiesoftheNPs.Themostobviousexampleofthischangeinpropertieswithsizeisquantumdots
which exhibit different colours on exposure to ultraviolet light depending on the size of the NPs.3, 4
MorphologycontrolstheNPpropertieshencebeingabletocontrolthesizedistributionandsizeofthe
synthesized NPs precisely is beneficial. Traditionally batch reactors have been used to explore new
synthetic routes to obtain a variety of different NP shapes such as spheres,5 plates,6 disks,7 rods,8
flowers,9urchins,10octopods,11wires12andcoreshellstructures.13SuchawidearrayofNPshapeshave
beenthesubjectofinterestbecauseofthechangingpropertieswhichachangeintheshapeoftheNP
wouldafford.
Themain focusof the researchpresentedhereinhasbeenonsilverandgoldNPs, the literature
presented hereafter focuses on these materials. Silver NPs are used in many applications, such as
sensors,14-16optics,17-19electronics,20, 21 catalysis,22 23, 24biomedicine,25-27andsurfaceenhancedRaman
scattering.28 Gold NPs can be used in biological/biomedical,29, 30 electrochemical,31 catalytic,32 and
opticalapplications.19Withthevastamountofpotentialapplicationspresented,even justconsidering
goldandsilverNPs,thereisthenaneedtobeabletocontrolthesizeanddispersityofNPsforthemto
display suitableproperties forany intendedapplication.Besides this, it is alsodesirable tobeable to
produce NPs in suitable quantities in a reliable manner. Flow reactors have the advantage over
traditionally usedbatch reactors for the synthesis ofNPs, because they are less labour intensive and
Chapter1
19
more conducive to mass production of materials. Given that the size and dispersity of the NPs
determines their properties and hence application, it is important to consider the advantages and
disadvantagesofthecharacterisationtechniquesforNPanalysisusedwithinthiswork.
1.3 Characterisationtechniquestodeterminethemorphologyofnanoparticles
Thefollowingsectioncontainsabriefdescriptionalongwiththeadvantagesanddisadvantagesofthe
characterisation techniquesused in thiswork todetermine the size anddispersity of the synthesised
NPs.ThecharacterisationtechniquesusedwereUV-Visspectroscopy,DCSandTEM.
1.3.1 Ultraviolet-visiblespectroscopy
UV-Vis spectroscopy allows the measure of the interaction of NPs with the visible and ultraviolet
portions of the electromagnetic spectrum. This interaction is highly dependent on thematerial, size,
shapeanddispersityoftheNPsbeingmeasured.SmallchangesinthesizeoftheNPscancausechanges
intheirinteractionwithlightwaves.
1.3.2 Differentialcentrifugalsedimentation
DCSusescentrifugal force toseparateparticlesofdifferentsizebasedon theirdensity.SinceNPsare
very small, they do not settle in aqueous solutions very quickly and therefore a centrifugal force is
appliedusingaspinningdisctoallowsmallerparticlestosediment inareasonabletimeframe.Stokes’
Law is used to determine the distribution of sizes in a nanoparticle population by determining the
amountoftimerequiredforagivensizeofparticletosettleinasolutionofknownviscosityanddensity.
Theoperationof theDCS involves the injectionof a sampleofNPs into the centreof a spinningdisk
whichcontainsasolutionofsucroseinwaterwithadensitygradient.ThesizeoftheNPdeterminesthe
speedatwhichitreachesthelightsourceanddetector.Theintensityofthelightisrecordedovertime
and changes depending on the concentration of NPs passing through the light source and detector
region.ThetimecanberelatedtosizeusingStokes’LawandinthiswaythesizedistributionoftheNPs
canbedetermined.
1.3.3 Transmissionelectronmicroscopy
The TEM operation can briefly be described as follows. An electron gun generates electrons which
travel through a vacuum generated within the microscope. The electrons pass through an
electromagneticlenswhichfocusesthebeam.Thistravelsthroughthespecimen(athincarboncoated
coppergridwhichtheNPsampleisplacedon),wheretheNPsinteractwiththebeamandelectronsare
absorbedorscattered.Thebeamtravels throughthespecimenandthroughvarious lenses toreacha
detector (such as a fluorescent film)where the sample is visualised. Areaswhere the electron beam
interacted stronglywith the sampleandwereabsorbedand/or scatteredappeardark,whereasareas
wheretheelectronsdidnotinteractappearlight.Inthisway,theNPsarevisualised.
Introduction
20
1.4 Microfluidics
Microfluidicsisabranchofflowprocessesconcerningflowreactorswhichhavereactionchannelswith
diameters in themicrometre range (<1mm).Thishasbeenanactiveareaof research in recentyears,
uncoveringmanychallengesandbenefitsthatmicrofluidicspresent.33
In this study, various types of microfluidic devices have been employed to synthesize gold and
silverNPswiththeaimof investigatinghowdifferentdesignscanbeusedtomanipulatetheresultant
size and size distribution of the NPs. The size and dispersity of the NPs are largely controlled by a
chemicalreactionandsubsequentnucleationandgrowth.34Theseareinturncontrolledbyparameters
suchasconcentrationofreagents,pHandtemperature.Usually,achangeinmorphologicalparameters
is brought about by changing one of the above parameters. Currently,many studies in the literature
focusondoingexactlythis.Thisapproachenablesavarietyofshapes,sizesandcompositionsofNPsto
beobtainedinarelativelysimplemanner.Ontheotherhand,masstransferwithinthereactorhasbeen
thefocusofmuchfewerstudies.Themasstransfergovernstheconcentrationsatwhichthereactions
takeplace,andmicrofluidicdevicesoffertightercontrolovermasstransferwhencomparedtoalarger
volume flow or batch reactor. Another benefit of using microfluidic devices is the spatiotemporal
separationofthevariousstagesinthesynthesisincludingreaction,nucleationandgrowthwhichallows
foragreatercontroloversize,shape,compositionandsizedistributionoftheNPs.35-37Italsoallowsfor
anincreasedflexibilityinanalysingthemechanismsofeachofthesestages.38-40
Asmentionedpreviously,themajorityofresearchsynthesisingNPshasbeencarriedoutinbatch
processes,e.g.bringingthereactantstogetherandallowingthemtoreacttoformproduct,followedby
removaloftheproductsaftersomereactiontime.Themaindifferencebetweenabatchprocessanda
flow process reaction is that product is continuously withdrawn as reactants are pumped in a flow
process. Figure1-1 showsa schematicofabatch reactorandFigure1-2 showsa schematicofa flow
reactor.
Figure1-1:Schematicrepresentationofabatchreactor
Chapter1
21
Figure1-2:Schematicrepresentationofaflowreactor
AkeyconceptinmixingisthedimensionlessReynoldsnumber,describingtheratioofinertialforcesto
viscousforcesinafluid:
Re=inertialforcesviscousforces
=𝜌ud𝜇
where𝜌isthedensityofthefluid,uisthevelocityofthefluid,disthediameterofthechanneland𝜇is
theviscosityofthefluid.Thisdimensionlessnumber isuseful incharacterisingthetypeofflowwithin
thesystem.Therearetwodistinctcharacteristicflows,laminarflowandturbulentflow.Laminarflows
occur at lowReynolds numbers, typically at Re < 2000. Turbulent flow typically occurs at Re > 2000
(though in amicrochannel, turbulence can occur at lower Reynolds numbers). There is a transitional
phasebetweenthetwocharacteristicflows.Inthelaminarregime,mixingoccursthroughdiffusionand
thevelocityprofileisnon-uniformacrossthechannel(sloweratthewalls,fasterinthecentre).Inthe
turbulent regime, non-uniform velocity profiles form resulting in a stretching and thinning of fluid
lamellae. This improves the macromixing (time scale of bulk homogenisation) and also reduces the
diffusion distance as fluid lamellae are stretched resulting in faster mixing on the microscale. A
schematicrepresentationofthesetypesofflowwithinachannelisshowninFigure1-3.
Figure1-3:Schematicrepresentationofturbulentflowandlaminarflow
Introduction
22
Withinamicrochannel,Reynoldsnumbersareusuallyverylow(Re<100)becauseofthesmallchannel
diameterthefluidflowsthrough.Thismeansthatflowgenerallyfallsintothelaminarregimesomixing
primarilyoccursthroughdiffusion.Theperformanceofamixercanbemeasuredbytheamountoftime
ittakestomixreactantstogether(mixingtime).Mixingtimecanbeseenastheamountoftimeittakes
forreactantstoachieveahomogeneousconcentrationprofile.Dependingonthedegreeofturbulence
inthemicrochannel,mixingwilloccurthroughamixtureofchaoticadvectionanddiffusion.Diffusional
transport is described by Fick’s Law, and the characteristic mixing time can be deduced from this
relationshipasfollows:
t∝dt2
D
wheretisthemixingtime,dtisthelamellawidthandDisthediffusioncoefficient.33Itcanbeseenfrom
this relationship that mixing time is dependent on the distance across which a given species has to
diffuse tomixwith another species. Thus, to achieve quickmixing times it is necessary to efficiently
reduce the diffusion distance between different specieswhich are to bemixed.Ottino describes the
striationthickness,s,as:
s=12(sa+sb)
wheresaandsbarethicknessesofstreamAandstreamBrespectively,asthesimplestmeasureofthe
stateofmechanicalmixing.41Thisislinkedtothemixingtimebeingproportionaltothelamellawidthas
stated by Fick’s Law.Mixing inmicrofluidic devices is generally focused on this reduction of striation
thickness to improve mixing efficiency. Mixing efficiency is a key parameter when considering mass
transfer.Therearesomeotherpertinentdimensionlessnumbers,besidestheReynoldsnumber,when
consideringmixingwithinthemicrofluidicdevicesstudiedinthefollowingchapters.ThePecletnumber
isusedtodefinethedominantmechanismofmasstransportinafluidandisaratioofthetimescalesof
diffusionandadvection:
Pe=diffusiontimeadvectiontime
=( L
2
D )
( uL )=
LuD
whereListhecharacteristiclength,uistheflowvelocityandDisthediffusioncoefficient.AhighPeclet
number (>> 1) is indicative of an advection dominated mass transport mechanism whereas if the
number is low (<<1) it is an indicativeof a diffusiondominatedmass transportmechanism.Another
importantnumber is theWebernumber,which is a ratiobetween the inertial forcesand the surface
tensionforcesinafluid:
Chapter1
23
We=ρu2dσ
whereρ isthedensityoffluid,uisaveragevelocityoffluidbeforeimpact,disthediameterofthejet
(takenas the tube internal diameter) andσ is the surface tensionof the fluid. TheWebernumber is
useful when describing the behaviour of impinging jets (chapter 4), as it is useful in characterising
whetherthereisacontinuousjetafterimpingementorthereisdropletbreakupawayfromthejetafter
impingement.Thisisanimportantconsiderationwhenlookingatthemixingefficiencyofimpingingjets.
1.5 Mixing
Mixingoccursprimarilythroughtwomechanisms,advectionanddiffusion,whichcanbedescribedina
more encompassing term, convection. Diffusion describes themovement ofmolecules from areas of
highconcentrationtoareasof lowconcentrationand isrelativelyslow.Whenconsideringmicrofluidic
devices,theprimarymodeofmasstransport isthroughdiffusion(becausetheflowisgenerally inthe
laminar regime). However, the distance across which diffusion of molecules occurs in a microfluidic
device is small. Usually this distance is reduced further by manipulating the flow in themicrofluidic
devicestocreatethinnerlamellaeoffluid(differentmixingdesignsarediscussedinchapter2).Laminar
flow is a result of the viscous forces inmoving fluid dominating the inertial forces (described by the
Reynoldsnumber),resultinginasmoothflowwithstraightstreamlinesalongthedirectionoftheflow.
When the inertial forcesbegin to increase, eitherby increasing the fluid velocityor applyingexternal
force such as stirring of the fluid, the streamlines no longer maintain a smooth streamline in the
directionoftheflow.Themovementofthefluidcanthenenhancethemixingthroughadvection,where
thefluidelementsarestretchedandfolded,ultimatelyleadingtofastermixingbecausediffusioncanact
overashorterdistance.
1.5.1 Diffusion
Fick’s law of diffusion describes the flux of a species soluble in a fluid, where the specieswill travel
withinthefluidfromanareaofhighconcentrationtoanareaoflowconcentration.Consideringspecies
𝑎travelingthroughmedium𝑏inthe𝑦direction,thelawcanbestatedas:
𝑛)* = −𝐷).𝑑𝐶)𝑑𝑦
where𝑛 is the flux [mol/m2s],𝐷). is themolecular diffusivity of species a inmediumb and𝐶 is the
concentration [mol/m3]. Diffusion describes the movement of material at the smallest scale, and is
essentiallyhowallmixingoccurs.Thespeedatwhichthisprocessoccursisdependentonthediffusivity
of thematerial and the difference in concentration between the areas thematerial has to travel (as
describedbyFick’sLaw).
Introduction
24
1.5.2 Advection
Advection of a fluid enhances themass transfer through themotion of the bulk fluid, which can be
describedby its velocityprofile.Velocityprofileswithina fluid canbecomputedbyusing theNavier-
Stokes Equations. This enhancement in mass transfer occurs when fluid moves in such a way to
experience increased exposure to the surrounding bulk fluid. In laminar flow, each fluid element has
limitedinteractionwiththebulkfluid,becausethevelocityprofileconsistsofstraightstreamlineswhich
causethefluidelementstostayona linearpathwhichcannottravelthroughoutthebulkfluid.When
the flow transitions into a turbulent flow, the fluid elements followmore complex streamlineswhich
allow them to interactwithmore fluidelements in thebulk fluid, aswell as causinga stretchingand
thinning of these fluid elements to reduce diffusion distance, enhancing themass transport through
diffusion.
1.6 Objectives
Synthesis of NPs consists of reaction, nucleation and growth kinetics, also known as “reactive
crystallization”. Changes in size and dispersity of NPs can be achieved by changing reagent
concentrations,thusaffectingkineticsoftheprocessesinvolvedinreactivecrystallization.Thisapproach
isoftenusedfromachemistryperspective.Theaimofthisstudy is to investigatehowchangingmass
transfer conditions affect NP size and dispersity from a chemical engineering perspective, hence
considering flow hydrodynamics, mixing efficiency and concentration profiles within the microfluidic
devices.Thisisachievedbyusingvariouscharacterizationtechniquesonthemicrofluidicdevicessuchas
mixing time characterization using a competing parallel reaction scheme, high speed camera imaging
andflowvisualizationusingamicroscope.Thischaracterizationenablesanunderstandingofthemass
transferineachofthemicrofluidicdevicesusedinthiswork.SynthesisofsilverandgoldNPscanthen
beperformedunderavarietyofdifferentmasstransferconditionsaffordedbythemicrofluidicdevices
used.ThisallowsaconnectiontobemadebetweenthesizeanddispersityofNPssynthesizedandthe
masstransferconditionstheyweresynthesizedunder.
1.7 Chapteroverview
§ Chapter 2 contains a literature review concerning microfluidic devices, silver NPs and gold NPs.
Various designs of microfluidic devices used to achieve ‘passive’ mixing were reviewed (‘active’
mixerswerenotconsideredbecauseoftheir increasedcomplexity).SilverandgoldNPsynthesis in
batchreactors,flowreactorsandanoverviewofthemechanisticstudiesonnucleationandgrowth
foreachmaterialarealsocovered.
§ Chapter3 containsa studyon the synthesisof silverNPsusinga coaxial flow reactor (CFR). Silver
nitrate is reducedviasodiumborohydride inthepresenceof trisodiumcitratewithintheCFR.The
effectof flowrate,precursorconcentrationand ligandconcentrationonNPsizeanddispersityare
presented.
Chapter1
25
§ Chapter4containsastudyonthesynthesisofsilverNPsusinganimpingingjetreactor(IJR).TheIJR
ischaracterizedusingtheVillermaux-Dushmanreactionsandhighspeedcameraimages.Detailson
the Villermaux-Dushman reactions and the interaction by exchange with the mean (IEM) mixing
modelarediscussed.SilverNPsaresynthesizedbyreductionofsilvernitrateviasodiumborohydride
witheither trisodiumcitrateorPVAusedasa ligand.Theeffectof flowrateonsilverNPsizeand
dispersityusingIJRsoftwodifferentjetdiametersispresented.
§ Chapter5containsastudyonthesynthesisofgoldandsilverNPsusingaCFR inconjunctionwith
othermicrofluidicdevices(coiledflowinverterandsplitandrecombinemixer)aswellasusingthe
CFR in isolation with different types of hydrodynamics (vortex flow and variation of stream
thickness). GoldNPs are synthesized by reduction of tetrachloroauric acid via trisodium citrate at
elevated temperatures, silver NPs are synthesized by reduction of silver nitrate via sodium
borohydrideinthepresenceoftrisodiumcitrate.Theeffectofvariationsinmasstransferconditions
andtemperatureonNPsizeanddispersityarepresented.
§ Chapter6containsastudyonsynthesisofgoldandsilverNPsusingabatchreactor.GoldNPsare
synthesized by reduction of tetrachloroauric acid via trisodium citrate at elevated temperatures,
silverNPsaresynthesizedbyreductionofsilvernitrateviasodiumborohydride in thepresenceof
trisodiumcitrate.Theeffectofmixingefficiencyand theorderof reagentadditiononNPsizeand
dispersityarepresented.
§ Chapter 7 summarizes conclusions drawn from the work presented along with possible future
directions.
26
2 LiteratureReview
2.1 Microfluidicdevicesformasstransfer
The following section gives an overview of microreactor technologies, in particular micromixers,
currently inuse.This istoaid intheunderstandingofhoweachtechnologyworks,thefocusbeingon
themethodofmixing.
There exist awide variety ofmicromixers reported in the literature. These are divided into two
categories, passive and active micromixers.42, 43 Active micromixers require the application of an
externalforcetoenhancethemixingperformance,andaremorecomplextofabricateandpackagein
general than passivemicromixers. Passivemicromixers rely on the geometricalmanipulation of fluid
elements within to achieve mixing.44 This section focuses on passive micromixers, a simpler type of
micromixerwithmanymorestudiesreportedintheliterature.
2.1.1 Parallellamination
T-mixers, 45-47 Y-mixers,48 interdigitated streams49 and hydrodynamic focusing fall50 into this class of
micromixing.Thebasicconceptbehindthemixingistoreducethestriationthicknessbetweenstreams
tobemixed,whichinturnreducesdiffusiondistanceandspeedsupmixing.Figure2-1isanexampleof
aninterdigitatedparallellaminationmicromixerstudiedbyBessothetal.49
Figure2-1:Exampleofaparallellaminationmixer.49ReproducedfromRef.49withpermissionofTheRoyalSociety
ofChemistry.
2.1.2 Seriallamination
This type of micromixer consists of a number of splitting and recombination (SAR) steps a stream
consistingoftwocomponentstoproduceprogressivelythinnerlamellaetodecreasethediffusionpath.
This form consists of a number of elements which will perform successive SAR of the streams. The
differencebetweenparallellaminationandthistypeisthatstreamsthataresplit,arerecombinedafter
onesplit.Thisisthenrepeatedtoproduceprogressivelythinnerlamellae.Acharacteristicadvantageof
this type of mixer is that the flowrate is not the main factor in determining mixing performance.
Schonfeld et al.,51 Kashid et al.,52 Buchegger et al.53 and Chen et al.54 have all investigated serial
laminatorsofdifferentdesigns.
Chapter2
27
2.1.3 Segmentedflow
This type ofmicromixer functions by splitting a continuous fluid flow into smaller discrete segments,
similar to tiny individual reactors flowing through a larger flow reactor. These smaller segments are
generally dispersed in a non-reactingmediumwhich carry the smaller segments through the reactor.
Themovementofthesesegmentscreatesaninternalflowfield,42thusmixingisgeneratedwithineach
segment. Waelchi and Von Rohr,55 Kashid et al.,56 Tanthapanichakoon et al.57 have investigated
segmentedflowreactors.Figure2-2givesavisualdepictionoftheinternalflowfieldsinslugstypically
foundinsegmentedflow.57
Figure2-2:Slugswithamixtureofbluedyeandwatershowingadvectionwithin.ReprintedfromRef.57,Copyright
(2006),withpermissionfromElsevier.
2.1.4 Injectionmicromixers
Thebasicmixingprinciplebehindinjectionmicromixersisonereagentisinjectedintoanotherreagent
streamthroughanumberofsmallnozzleswhichreducethestriationthicknessofthestreams,resulting
inquickerdiffusion.Miyakeetal.58,59andVoldmanetal.60haveinvestigatedinjectionmicromixers.
2.1.5 Chaoticadvectionmicromixers
Chaotic advection can be described as themovement of fluid in such away that the interfacial area
betweenreagents tobemixed iscontinuously increased,accompaniedbya reduction in thediffusion
length between reagents (striation thickness).61 Usually some kind of internal structures within the
microchannelareusedtoinduceflowfieldstoincreasetheinterfacialareabetweenfluidstobemixed.
Stroocketal.62andKangandKwon63havestudiedthestaggeredherringbonemixer(SHM),which isa
typeofchaoticadvectionmicromixer.
2.2 Silvernanoparticles
ThissectionpresentssilverNPsynthesisstudiesusingbatchreactors,microfluidicdevicesandthosethat
investigatemechanismsinsilverNPformation.ThemainfocusofthebatchreactorsectionissilverNPs
synthesizedusingsodiumborohydrideasareducingagentgiventhisisthesystemthathasbeenstudied
in thiswork.Awider variety of systemshavebeen reviewed for studiesusingmicrofluidic devices to
highlight thepossibilitiesofcontrollingsizeanddispersityof synthesizedNPs.Studies that investigate
themechanismsofreaction,growthandnucleationofsilverNPsarealsoreviewed.
LiteratureReview
28
2.2.1 Batchsynthesisofsilvernanoparticlesviasodiumborohydridereduction
AvarietyofreducingagentsexistforreductionofsilverprecursorstoformNPs.Inthissection,thefocus
isonsodiumborohydrideasareducingagentgiventhatthisthereagentusedinthiswork.
Creightonetal.usedicecoldsodiumborohydridetoreducesilvernitratetoproduceNPsinquitea
wide range between 1-50 nm with some sols produced being in the range between 1-10 nm.64
Interestingly,thesolswerestableoveraperiodofseveralweeksintheabsenceofanyligand.
Shirtcliffeetal.usedsilvernitrate,sodiumborohydrideandsodiumhydroxideasreagentsforthe
synthesisofsilverNPsintherangebetween15-77nm.65TheyusedEppendorfpipettestomixreagents
withinacuvetteat2°Candfoundthatthemethodandorderofmixinghadasignificanteffect.Addition
ofsilvernitratetoamixtureofsodiumborohydrideandsodiumhydroxideproducedthenarrowestand
mostreproduciblepeaksusingUV-Visanalysis(thoughtheTEMimagesshowedquitelargepolydisperse
NPs,possiblybecauseofthelackofligandusedinthesystem).
Ryuetal. synthesizedhighlyconcentratedsilverNPsamples (upto40wt%)byreducing icecold
silvernitrateviaslowadditionofsodiumborohydrideand/orhydrazinemonohydrateinthepresenceof
polyelectrolytesintheMWrangebetween1200-30000.66NPswereintherangebetween1-35nmwith
themajorityofconditionstestedproducinglessthan10nmNPs.
Songetal.produced silverNPsbyadding silvernitrate toa solutionof sodiumborohydrideand
sodiumdodecylsulfate.67Theyobservedthatincreasedsodiumborohydrideresultedinlessaggregation
oftheNPs,pointingtothehydrolysisofsodiumborohydridecausingaggregationbutwhenahighmolar
ratio(sodiumborohydridetoprecursor)wasused,athicklayerofBH4- ionspreventedboronhydroxide
from attaching onto the surface and destabilizing the NPs. However another possibility is that the
hydrolysis reaction causes an increase in pH which ultimately suppresses the hydrolysis reaction,
preventing the switching point which results in a sudden shift to a higher size due to coalescence,
becauseoftheconsumptionofsodiumborohydride.68
Pinto et al. synthesized silver NPs in the size range between 3-10 nm by addition of sodium
borohydridetoamixtureofsilvernitrateandtrisodiumcitrate.69Onincreasingsodiumborohydrideto
silvernitratemolarratios,thesizeoftheNPsdecreasedandtheconcentrationofNPsincreasedbased
onUV-Vis spectroscopy analysis. The standard deviation of the absorbancemaximum increasedwith
increasingsodiumborohydridemolaramounti.e.alowerreproducibilitybetweenexperiments.
Thøgersenetal.synthesizedsilverNPsinthesizerangebetween6-19nmbyadditionoficecold
AgNO3slowlyviadropstoan icecoldsolutionofsodiumborohydride.70Theyfoundan increasingsize
whenthemolarratioofsilvernitratetosodiumborohydridewasincreased.
Agnihotrietal.synthesizedsilverNPsintherangebetween5-100nmbyaddingsilvernitratetoa
heatedsolutionofsodiumborohydrideandtrisodiumcitrate(60°C),followedbyfurtherheatingofthe
solution to 90°C.5 They were able to synthesize monodispersed NPs and they suggested a two-step
growthmechanism,where sodiumborohydrideproduced small seeds througha coalescenceof small
silver clusters followed by a surface reduction type growth via citrate to grow the NPs into a more
monodispersed sample. They were able to synthesize a wide range of sizes by varying the
concentrationsof thereagentsused in thesystem. Ingeneral,alteringtheratiosof reducingagent to
Chapter2
29
precursorwasthemainmethodtoachievecontrolofsizeoftheNPs.Thishighlightstheimportanceof
concentrationintheresultantNPsize.
2.2.2 Microfluidicsynthesisofsilvernanoparticles
Wagneretal.haveusedsplitandrecombinemicromixersinserieswithresidenceloopsinbetweento
allow the growth of silver NPs.71 They reduced silver nitrate using sodium borohydride. Synthesized
silverNPshadameandiameterof15nm.Thestudyshowedadependencyofthesizedistributionofthe
NPs on the flowrate, with an increased flowrate resulting in a narrower size distribution. A faster
flowratewouldresultinbettermixing,withanarrowersizedistributionpointingtoasharperdistinction
between nucleation and growth periods. This suggests mixing is extremely important in obtaining
monodisperseNPs.Kohleretal.usedasimilarsplitandrecombinemixeraswellasanumberofother
configurations, which included the use of t-mixers and segmented flow techniques to synthesize
compositegoldandsilverNPs.72Tetrachloroauricacidandsilvernitratewereusedasprecursors,while
ascorbicacidwasusedasareducingagent.
Linetal.mixedsilverpentafluoropropionateinisomethyletherinthepresenceoftrioctylamineto
forma singlephase.73 SilverNPswere formed through thermaldecompositionusinga simple tubular
microchannel which was immersed in an oil bath held at temperature between 100-140°C. The
experimentalsetupisshowninFigure2-3.
Figure2-3:ExperimentalflowreactorsetupofsilverNPsynthesisusedbyLinetal.Reprintedwithpermissionfrom
Ref.73.Copyright(2004)AmericanChemicalSociety.
Synthesizedparticlesrangedfrom7.4to8.7nminsizewithsizevariancerangingfrom10%to19%.The
study showedan increase in flowrate resulted ina smalleraveragediameterof theNPs, though they
weremorepolydisperse.Thestudysuggeststhisispossiblydowntowalleffectsplayingarole,resulting
inaparabolicvelocityprofilewithinthetube.Sincethissystemisonephase(reagentsarepremixed),
theoreticallymixingshouldnotbeanissue(assumingtheyarewellmixedbeforehand).Theactivationof
thereactionoccursthroughan increase intemperature.Therefore,toobtainmonodispersesilverNPs
the temperature profile must be uniform. Since microreactors have extremely high surface area to
volume ratios, they excel at heat transfer. Assuming the temperature profile is uniform, the other
possibilitythatleadstopolydispersityisabroadresidencetimedistribution(RTD).
Sonnichsenet al. used a simple setup involving a threeway valve followed by a residence loop
(tubing inner diameters of 0.75 mm for silver and 1mm for gold) shown in Figure 2-4.74 They
LiteratureReview
30
synthesised both gold and silver nanorods based on themethod by Jana et al.75 Though reasonable
resultswereobtainedforgoldnanorods,silvernanorodsweremoredifficult.Theresultsshowedawide
rangeofdifferent sizesand shapes inTEM images (Figure2-4). They reported residence timeswitha
spreadofabout12minuteswhentheyinjectedapulseofpre-madeNPsintothereactor.Thisleadsto
poorsizecontroloftheNPsbecauseresidencetimesdifferbetweendifferentNPsflowingthroughthe
reactor.Thiswouldresultinbroaderparticlesizedistributions.Similaraffectswereseenintheworkof
Linetal.73
Figure2-4:MicroflowsetuptosynthesisesilvernanorodsandTEMimageofsilvernanorods.ReproducedfromRef.
74withpermissionofTheRoyalSocietyofChemistry.
Knaueretal.employedamicrosegmentedflowtechniquewithinamicrochanneltosynthesisedouble
shellNPswithagoldcorewhichhasasilvershellandasubsequentgoldshell.76Tetrachloroauricacid
andsilvernitratewerereducedbyascorbicacidinthepresenceofcetyltrimethylammoniumbromide.A
narrowersizedistributionforthesynthesizedNPswasobtainedinthecaseofthemicrosegmentedflow
reactor in comparison to the batch synthesis. This supports the idea of better control over size
distributionswithabetterRTD.AnarrowRTDisacharacteristicofsegmentedflowtechniques.77
Koehleretal.alsosynthesizedAu/AgcompositeNPsusinganumberofsplitandrecombinemixers
in series to bring the reactants together.78 Tetrachloroauric acid and silver nitrate were reduced by
ascorbicacid.Theimportanceofmixingwashighlightedinthisstudy,indicatedbythediverserangeof
differentNPcompositesobtainedbychangingmixingorders,mixing timesand flow rates (whichalso
affectstherateofmixing).
Liuet al. studied the effects of various parameters on the size distribution of synthesized silver
NPs.79 The synthesis is a thermal reductionof silver nitrate using plant extract fromC.Platycladiwith
additionofsodiumhydroxide.ThesetupisshowninFigure2-5.
Chapter2
31
Figure2-5:ExperimentalsetupofmicroreactorusedbyLiuetal.ReprintedfromRef.79,Copyright(2012),with
permissionfromElsevier.
Particlesizeincreasedwhenflowratewasincreasedfrom0.2ml/minto0.4ml/min(3.5nmto7.2nm)
andthendecreasedwhentheflowratewasfurtherincreasedfrom0.4ml/minto0.6ml/min(7.2nmto
3nm).IncreasingC.Platycladiconcentrationresultedinincreasingparticlesizeanddispersity(4.7±1.9
nmto15±8.6nm).Themicroreactorpropertiesalsoplayedakeyrole.T-typemixersproducedsmaller
andmorenarrowlydistributedparticlesthanthatproducedbyY-typemixers(4.7±1.9nmagainst5.3±
2.9nmrespectively).Particlesproduced inPTFE reactorsweresmallerandmorenarrowlydistributed
than those produced in ETFE (4.2 ± 0.9 nm against 4.7 ± 1.9 nm respectively). Both reactors had a
channel diameter of 0.8 mm. Smaller channel diameters (0.5 mm, 7.0 ± 1.9 nm) produced larger
particles on average than larger channel sizes (0.8 mm, 4.7 ± 1.9 nm), however the flowrate was
changedtoproducesimilarresidencetimeswhichmayhavealsoaffectedflowhydrodynamics.
Kumaret al. synthesised silverNPs in a spiral flow reactorusing segmented flow techniques via
inert liquid kerosene and inert gas air.80 Stearic acid sophorolipid molecules were used as both a
reducingandcappingagent.Silvernitratewasreducedbythesemoleculesat90°C.Figure2-6showsthe
setupusedinthestudy.
LiteratureReview
32
Figure2-6:Spiral flowreactorusedbyKumaretal., 80Wrepresentsreactingphase,Krepresents liquidkerosene
inertphase,A representsgasair inertphase.B represents inletmicromixer (crossmixer)used togenerate slugs,
withPEEKandStainlessSteelbeingused(dimensionsgiven).CandDarePMMA(0.5mmID)andStainlessSteel(1
mmID)spiralreactorsrespectively.ReprintedfromRef.80,Copyright(2012),withpermissionfromElsevier.
They studied the effects of variations in the slug flow (liquid-liquid, gas-liquid, size), flow rate, inlet
micromixerdiameterandflowrateratioshaveonthesizedistributionoftheformedsilverNPs.Various
effectswerereported,themainonebeingthatsmallerandnarrowersizedistributionswereobtained
whenthereactivephasewascontinuousratherthaninslugform.Itissuggestedthatinternalcirculation
(mixing),flowvelocitiesanddifferencesinmaterialproperties(liquid-liquidandgas-liquid)contributeto
theparticlesizedistribution.Flowratesusedwere35μl/min,100μl/minand1ml/minwithratiosof
inert phase to reacting phase being 0.5, 1 and 2. At lower flowrates, smaller and more narrowly
distributedparticleswereformed,withair/waterat35μl/min(2:1airtowater)performingthebestin
termsofnarrowdistributionwiththesmallestsize.
Lazarusetal.studiedthesynthesisofsilverandgoldNPsusingsegmenteddropletflowusingionic
liquid.81 Droplet formation of viscous BMIM-Tf2N (1-butyl-3-methylimidazoliumbis-
(trifluoromethylsulfonyl)imide)withinacontinuousfluorocarbonoilphase,polychlorotrifluoroethylene
(PCTFE), was achieved by coating the PDMS microfluidic device with a fluoropolymer via initiated
chemicalvapourdeposition(iCVD).Thisallowsthecarrieroilphasetocoatthesurfaceofthechannel
walls to prevent the ionic droplets from contacting the wall (improving the residence time because
segmentedflowisknowntoimproveRTD,butfurtherimprovementswouldbeobtainedifwalleffects
areminimisedasinthiscase).Dropletformationatanumberofdifferentflowrateratioscanbeseenin
Figure2-7.
Chapter2
33
Figure2-7:Configurationofmicrofludicdevicewithfour inlets:1)Carrierfluidinlet,2)Precursorsolutioninlet,3)
Solventinlet,4)Reducingagentinlet.Righthandsideshowssegmentedflowofionicliquid(BMIM-Tf2N)withinthe
oil carrier phase (PCTFE).At higher flowrates, jettingoccurs further along the channel as thedroplets pinch and
formthinstrandphases.ReprintedwithpermissionfromRef.81.Copyright(2012)AmericanChemicalSociety.
In the study, 1-butyl-3-methylimidazolium borohydride (BMIM-BH4) is used as a reducing agent
preferentiallyover sodiumborohydride.1-methylimidazolewasusedasa stabilisingagent,because it
binds tometal surfacesand servesas anacid scavengerwhichwouldprevent thedestabilisationand
agglomerationofNPsinthecaseofacidbuild-up.AgBF4wasusedasaprecursor(HAuCl4wasusedfor
gold).
Figure2-8:SilverNPsmadeusingA)Microfluidic device, B) using abatch reactor, C)UV-Vis comparisonof both
(microfluidic device is black, batch is red). Reprinted with permission from Ref. 81. Copyright (2012) American
ChemicalSociety.
Figure 2-8 shows the Ag NPs synthesised in the study with a marked difference between the two
synthesismethodsemployed,batchreactorandmicroreactor.ThemeansizeoftheAgNPssynthesised
in the microfluidic device are 3.73 ± 0.77 nm whereas those in batch produced large coral like
assemblies.TheUV-VisdatapresentedsuggestsahighnumberofsmallerAgNPsoflessthan10atoms
presentinthebatchsynthesis.
LiteratureReview
34
Overall,avarietyofdifferentmicrofluidicdeviceshavebeenusedforthesynthesisofsilverNPsto
showcasethepotentialbenefitstheypresent.
2.2.3 Mechanisticstudiesofsilvernanoparticlesynthesis
Polteetal. studied themechanismsof silverNP synthesisusinga continuous SAXS (small angle x-ray
scattering)setup.68Silverperchloratewasusedasaprecursor,sodiumborohydrideasareducingagent
andPVPasastabilisingagent.System1usednostabilisingagentandsystem2showsthesynthesiswith
PVPas a stabiliser, showing similar trendsbut taking longer than system1 in termsof thegrowthof
silverNPs(Figure2-9).Itseemsthatsurfactantsmayslowthekineticsdown(PVPinthiscase).Inboth
casesthegrowthmechanismisthesame,however,thehydrolysisofNaBH4(whichissupposedlysped
upbythepresenceofmetalNPs)slowsdownwiththeadditionofPVPresultinginalongertimebefore
the switching point. The switching point is the point at which particle stability decreases massively,
resultinginparticlescoalescingfrom2-3nmupto6-8nminsize.Thehydrolysisofsodiumborohydride
resultsinachangeinthecolloidalstability,resultinginaninevitableincreaseinthesizeofNPsdueto
coalescence coupledwitha reduction inpolydispersity.An illustrationof the suggestedmechanism is
showninFigure2-9.
Figure2-9:SuggestedgrowthmechanismforsilverNPsynthesis.ReprintedwithpermissionfromRef.68.Copyright
(2012)AmericanChemicalSociety.
Takesueetal.performedmechanisticstudiesonsilverNPsynthesisusingsilvernitrateasaprecursor
withcitricacidasastabiliserandtannicacidasareducingagent.82SAXSwasusedtocharacterisethe
NPsonlineinaflowreactorsetup(Y-mixer).Themoleratiosofsilvernitrate:citricacid:tannicacidwas
1:5:0.02withsodiumhydroxideusedtoadjustthepHto12.Themaingrowthmechanismsuggestedby
thisstudyisthatAg13clustersareformedwhichthencoalesceandaggregatetoformlargerNPs.Thisis
evidencedbyaninitialformationofseedparticleswhichmatchthesizeofanAg13cluster,followedby
gradual increase in the sizeof theNPs toa final average sizeofaround7nm.Ag55 clusterswerenot
formedinsignificantnumbersbecausethesizeofthisNPwasonlypresentinverysmallamounts.The
proposedmechanismcanbeseeninFigure2-10.Theonsetofnucleationbecomesclearat0.59msin
Chapter2
35
thissynthesis.0.71nmAgclustersformthemajoritybetween0.59and0.79ms,correspondingtoAg13
clusters,howeversomeparticleslargerthan10nmexistandtheprocessofgrowthisessentiallyoverat
3.93ms.TheauthorsuggeststheprocessofgrowthisdictatedbyAg13particlescoalescingandgrowing
toformlargerparticles.
Figure2-10:MechanismofsilverNPsynthesisproposedbyTakesueetal.82Silvernanoparticlesareformedthrough
(I)aninductionperiodduringwhichAg+ionsarereducedtoAg0atoms;(II)anucleation-dominantformationperiod
duringwhichAg13clustersarenucleatedandsimultaneouslyconsumedinsilvernanoparticleformation;and(III)a
growthperiodduringwhichlargersilvernanoparticlesareformedbycoalescenceandaggregationofsmallersilver
nanoparticles.ReprintedwithpermissionfromRef.82.Copyright(2011)AmericanChemicalSociety.
HengleinandGiersigsynthesisedsilverNPsusingAgClO4solutionasaprecursorwhichalsocontains2-
propanol,N2Oandsodiumcitrateinvariousconcentrations.83Thereducingagentisthe1-hydroxyalkyl
radicalgeneratedthroughradiolysis.Themainfindingsofthepaperisthatatlowcitrateconcentrations
theNPssynthesisedarepartlyagglomeratedwithmanyimperfections,atanintermediateconcentration
theparticlesarewellformedandseparatedwithnarrowsizedistributionsandathigherconcentrations
thereiscoalescenceofparticlesduetotheinstabilitycausedbythehighionicstrengthofthesolution.
Themolar ratioof sodiumcitrate toprecursor in the study ranges from0.5 to15. Theypropose two
mechanisms of silver NP growth (Figure 2-11). Surface reduction growth which occurs via electron
transfer, reducing Ag+ ions onto the surface of the particles already present in the solution, or
coalescenceofsmallerNPstoformlargerNPs.Thestudysuggeststhatifthecitrateconcentrationistoo
low,thesilverNPsarenotsufficientlycoveredwithstabilisingcitrateandcoalescenceoccurs.Athigher
concentrations, the ionic strength of the solution destabilises particles. A very important growth
experimenttheycarriedoutwhichinvolvedtakingaseedsolutionofsilverNPsandaddingfurthersilver
precursorandreducingelectronstocheckhowtheseedparticleswereaffected.Theexperimentshows
LiteratureReview
36
aperfect surface reductionoccurringas all the silver reduces tomakeexistingparticlesbigger rather
than produce new particles. This occurs by introducing 5 ml of seed particles to another 35 ml of
solution,whichafterirradiationproducesparticleswhichhavegrownperfectlyinsizeaccordingtothe
amountofsilveradded.This is importantbecause itshowsthepotential fortuningNPsizesbutmore
importantlythatthemechanismofsurfacereductionforgrowthofsilverNPsdoesexist.
Figure2-11:GrowthmechanismsproposedbyHengleinandGiersigatlowandoptimalcitrateconcentrationswith
twodifferentmechanismsoccurring. Atoptimalconcentrations, it is suggested thata surface reductionof silver
ionsoccursonthesurfaceofformedNPsowingtothereductionpotentialofthecitrateadsorbedontothesurface.
At low concentrations, coalescence of smaller silver NP clusters occurs.34, 83 Ref. 34 published by The American
ChemicalSociety.
PillaiandKamatstudiedtheroleof thecitrate ion in reducingsilver ionsaswellas its stabilisationof
silverNPsandtheformationofcomplexeswithsilverions.84Theyreducedsilvernitrateusingcitrateat
elevatedtemperaturesaswellasγradiolysis.Usingcitrateasareducingagent,ahigherconcentration
allowed the reaction to proceed faster as well as resulting in increased absorbance at 420 nm
wavelengthon theUV-Vispeak.This is inagreementwitha stronger reducingeffectproducing faster
reactions,aswellasproducingamoreconcentratedsampleat420nm,indicatingmoreNPsinthatsize
rangeweresuccessfullyformedandstabilised.Inthecaseofusingγradiolysis,theauthorsuggeststhe
silvermaynotreactat lowcitrateconcentrationsfullybecauseofa lowerpeakabsorbance.However,
the peak absorbance may be lower because of the presence of aggregates; naturally the peak
absorptionwillbelowerifthereisahigherconcentrationofaggregateswhichcontributeabsorbanceto
longerwavelengths(indicatedbymoreprominentshoulderpeaks).Citrateconcentrationinthecaseof
γradiolysiswillaffectthestabilisationoftheNPsandastheconcentrationincreases,theshoulderpeak
absorbancedecreasesindicatingdecreasingconcentrationofaggregates(Figure2-12).Higherreducing
power points to smaller average particle size from the difference in size between the twomethods
Chapter2
37
(Turkevitch,50-100nmandγ radiolysis,5-10nm),where thecitratehas less reducingpower than the
radiolysis(resultinginalowersupersaturation).Agrowthdominatedmechanismisfavouredwhenthe
reducingpower is relatively low (citrate) resulting in fewernucleiwhichgrow to larger sizesandvice
versaforhigherrelativereducingpower(γradiolysis).Thegrowthmechanismtheysuggestisshownin
Figure2-13.
Figure 2-12: Absorption spectra of Ag colloids prepared by γ irradiation of an aqueous solution (N2 purged)
containing 1 mM AgNO3, 1% methanol, and (a) 1mM, (b) 5mM, and (c) 10 mM sodium citrate as stabilizer.84
ReprintedwithpermissionfromRef.84.Copyright(2003)AmericanChemicalSociety.
Figure 2-13: Growth mechanism through pulse radiolysis of a mixture of silver nitrate and trisodium citrate
proposed by Pillai and Kamat: primary and secondary growth steps in the formation of silver nanocrystallites.84
ReprintedwithpermissionfromRef.84.Copyright(2003)AmericanChemicalSociety.
Zhangetal.synthesisedsilverNPsandnanorodsusingsilvernitrateasaprecursor,sodiumborohydride
asareducingagentandtrisodiumcitrateasstabilizingagent.85Thoughthisisnotaspecificmechanistic
study, there is useful information regarding mechanisms of silver NP formation. To obtain different
shapesofNPs,preferentialgrowthisrequired incertaindirections.Theevidence inthisstudysuggest
that nanocrystals formedwith little defects show a uniform coverage of citrate, whereas thosewith
defects are more suitable for nanorod growth because of the preferential adsorption of citrate on
different surfacesbecauseofdefects in the crystal structure. Thishas implicationsonmonodispersity
andshapecontrolofNPs.Moreover,thedifferenceindefectsofNPsformedarisesfromadifferencein
themixingofthereagents(allatonceanddropwise).
Woehletal.studiedtheeffectsofreducingpower(i.e.numberofelectronsavailableforreduction)
usingin-situTEMscanningonthenucleationandgrowthcharacteristicsofsilverNPs.86Reductionwas
achieveddirectlyfromelectronsacquiredfromtheTEMbeam,whicharetransferredtothesolutionand
reduce the silver nitrate. The authors use classical nucleation theory to explain the effects observed
LiteratureReview
38
pointing to a certain concentration of electrons (reducing power) needed to achieve the conditions
which induce nucleation (supersaturation) and observed there was no nucleation below a certain
threshold of electron concentration. They find at a dose rate much higher than this threshold, the
growthoftheNPsisdiffusionlimitedandiscontrolledbyhowquicklysilveratomscanfindthesurface
of the crystals from the bulk of the solution. In this case themorphology of crystalswas seen to be
spherical. At an electron dosage closer to that of the threshold for inducing nucleation, growth is
reaction limited in the sense that the surface reduction of silver atoms onto the NP is the slowest
process resulting inNPs that grow intodifferent shapes. Thesedifferencesareattributed todiffusion
limitationsinthecaseofsphericalparticles,wherethereisanabundanceofreducedsilverionswhich
can attach onto the surface of formed nanocrystals, and reaction limited in the non-spherical case
where there is a lack of electrons available near the surface of the particles leading to a surface
reductionreactionlimitingstep.ThesummaryoftheirinvestigationcanbeseeninFigure2-14.
Figure2-14:MechanismsofsilverNPgrowthusingdifferentbeamcurrentsunderaTEMmicroscope.86Reprinted
withpermissionfromRef.86.Copyright(2012)AmericanChemicalSociety.
Thevariousmechanistic studies in the literaturehaveperformed thesilverNPsynthesisusing slightly
different conditions (reaction vessels, chemicals, concentrations etc.), so must be considered when
makingcomparisons.ThestudiesbyPolteetal.andTakesueetal.bothuseSAXSandachievedsimilar
sizedNPs. 68, 82 They are in agreement that the overarchingmechanismof the silverNP formation is
coalescencewithnomentionofsurfacereduction(surfacereductionimpliedbythestudybyHenglein
andGiersig).83However,becauseofthelimitedconditionsatwhichtheSAXSstudiesarecarriedout,itis
possible that theremay be differentmechanisms depending on the conditions. Henglein and Giersig
andZhangetal.discussthepresenceofdefectsincrystals,whichissomewhatcorrelatedinthestudies
to differences in shapes and size distributions.83, 85 Different growth mechanisms depending on the
defectspresentonthecrystal surfacehavebeenhighlightedby thesestudies.Lastly,PillaiandKamat
and Jiang et al. discuss the complexes formed between citrate and silver ions,84, 87 which affect the
kineticsofthegrowth,addingmorecomplexitytothe issueofcontrollingthesize,distributionofsize
andshapeofthesilverNPs.
Chapter2
39
2.3 Goldnanoparticles
ThissectionpresentsgoldNPsynthesisstudiesusingbatchreactors,microfluidicdevicesandthosethat
investigatemechanisms ingoldNPformation.Themainfocusofthebatchreactorsection isgoldNPs
synthesized using the Turkevichmethod. Awider variety of systems have been reviewed for studies
usingmicrofluidic devices to synthesize gold NPs. Finally, studies that investigate themechanisms of
reaction,growthandnucleationofgoldNPsarereviewed.
2.3.1 Batchsynthesisofgoldnanoparticles
ThemostpopularmethodforsynthesisofgoldNPsisknownastheTurkevichmethod.Turkevichetal.
testedavarietyofdifferentmethodsbutfoundthesodiumcitratemethodtobethemostrepeatable.88
Thisfamousmethodinvolvesbringingastirredaqueoussolutionofchloroauricacidtotheboilbefore
addingasodiumcitratesolutiontotheprecursor.Subsequently,thecolourtransitionsthroughyellow,
clear,grey-blue,purpleandfinallyawinered.ThemethodproducesNPstypicallyintherangebetween
15-20nm.Thepopularityofthismethodisdowntoitseaseandreliabilityinproducingmonodispersed
gold NPs. The effect of temperature, sodium citrate concentration were studied by Turkevich et al.
Decreasing temperaturebelow100°Cresulted inslowingdownthereaction,butachange insizewas
observed.ThesmallestNPswereseenat80°Cbutincreasedslightlyinsizeifsynthesistemperaturewas
set to70°C.Reducing the sodiumcitrate concentration resulted in somemarkedchanges suchas the
colourchangesbeingmoredefinedwhile thegrey-bluewasmorebluewithdecreasedconcentration.
Theoverallsynthesiswasmorerapidatlowercitrateconcentrationswhiletheaveragediameterofthe
NPsincreased.Thecolourchangeofthesolution,fromtheinitialyellowofthechloroauricacidtothe
finalwinered,givesanideaofthemechanismstakingplacetoformtheNPs.
FrensworkedfurtheronthismethodandsynthesisandsynthesisedNPsintherangeof12-150nm
byvarying theamountofcitratewhichwas injected into theboilingsolutionof tetrachlorauricacid.89
Theconclusionof the studywas that the final sizeof theNPswasgovernedby thenumberofnuclei
formedwhichwascontrolledbytheamountofcitrateadded.Higheramountsofcitrateledtosmaller
NPs, and Goodman et al. observed a similar trendwith the same procedurewhen increasing citrate
concentrationwithNPsizefrom21to80nm.90
Chow and Zukoski synthesised gold NPs with sizes ranging between ca. 16-52 nm for
unagglomeratedparticles (very largeagglomeratedNPswereseen incaseswerethesynthesisofgold
NPswasnot complete).91Theyvaried temperaturebetween60-80°Cand found that the finalparticle
sizetendedtoca.20nmandittooklongertoarriveatthefinalsizewithdecreasingtemperature.
Suetal.synthesisedgoldNPsbysonicationoftetrachlorauricacidandtrisodiumcitratesolutions
at4°C.92Theyvariedthegoldprecursortocitratemolarratiofrom1:1to1:8andfoundthatnon-stable
largeaggregateswere formedat ratios atorbelow1:3andNPs in the size rangebetween17-22nm
wereformedatoraboveratiosof1:4.
Peietal.synthesisedgoldnanowireswithamodificationonthestandardmethodinwhich200ml
ofsodiumtetrachloroaurate isbroughtupto80°C inatemperaturecontrolledbathfollowedbyrapid
additionof trisodiumcitrate invarious controlledvolumes toadjust themolar ratioof citrate togold
LiteratureReview
40
from 0.1 to 2.7.93 Nanowireswere formed for ratios below 0.4, and particleswere formed for ratios
above0.4.Abluecolourwasassociatedwithwireformationwhichwasobservedatallratiosfrom0.2to
2.7,butthistransitionedintoaredcolourforratiosat0.4orhigherwhichisassociatedwithspherical
goldNPs.
Kimlingetal.synthesisedgoldNPsusingtheTurkevichmethod,UV-assistedreductionwithcitrate
and ascorbic acid reduction. The reaction volume for the Turkevich method was 100 ml, 2 ml in a
cuvettefortheUV-assistedreductionand50mlforascorbicacidreduction.94Theytestedgoldtocitrate
molarratiosofca.0.1-1.5forallthesyntheses.UsingtheTurkevichmethodtheysynthesisedgoldNPs
betweenca.17-130nmatconcentrationsoflessthan0.8mM,1mMand1.2mMtetrachlorauricacid.
They found that the size increasedwhen reducing the citrate concentration, attributed to the lackof
stabilizationofsmallerNPsbecauseof incompletecoverageofcitrate.Atconcentrations lessthan0.8
mM, the size of the NPs was independent of gold concentration and dependent on citrate
concentration. When gold concentration was increased above 0.8 mM the size of the NPs almost
doubledatlowercitratetogoldmolarratios.IncontrasttheUV-assistedreductionshowedasimilarsize
correlationforgoldconcentrationsof1and1.6mMwithNPsintheca.17-117nmrange.Theascorbic
acid reduction enabled synthesis of NPs between ca. 17-110 nm. This method showed an NP size
increasewhenincreasingthegoldconcentrationfrom0.5mMto1.6mMsimilartothatintheTurkevich
method,butwhenincreasingthepHto7,thesizeoftheNPswassimilarforbothgoldconcentrations
supposedlybecauseofareductionininstabilitiescausedbytheacidityatthehigherconcentrations.
Jietal. investigatedtheroleofcitrateinthestandardmethod.95Byvaryingthetrisodiumcitrate
concentration, they foundaminimumsizeofca.15nmtoexistatamolarratioof3.5to1trisodium
citratetotetrachloroauricacid, if theratiowas lowerorhigherthanthis therewas increase insizeof
NPs. They found that at higher citrate ratios, the reaction took longer to proceed. Nanowire like
structureswereobservedintheearlystagesofsynthesisatcitrateratiosinthelowerrange(<3.5)while
athigherratiostheNPswheresphericalinallstagesofthesynthesis.Theconsumptionofprecursorwas
foundtobeingeneralmorerapidatlowercitrateratios,inlinewiththeideathatreactivityandhence
nucleationrateishigheratlowerpH.BytuningthepHbutkeepingthecitratetogoldprecursormolar
ratio constant, the size was varied from small to large in the same way as increasing citrate
concentration would but the dispersity was reduced and more constant through the pH variation
methodfrom20-40nmNPsize.
Patungwasa and Hodak investigated the effect of pH on gold NPs.96 They found that NP size
decreasedwithincreasingpHintherangebetween4and6.5.Thiswassuggestedtobebecauseofthe
chargeonthecitrateadsorbedontotheNPsurface,atlowerpHthereisdecreasedchargeonthecitrate
surfaceresultinginreducedstabilitywhereasathigherpHthecitratehasahigherchargewhichwould
explain why smaller and less polydisperse NPs are seen at higher pH. The system tends to higher
averageNPsizeatlowpH.
Ojea-Jiménezetal.synthesisedgoldNPsusingthestandardmethodaswellasamodificationusing
deuteriumoxideasasolventinsteadofwater.97TheysynthesisedgoldNPsbyheatingtrisodiumcitrate
in either 100% H2O, 50% D2O/H2O or 100% D2O to which they added tetrachloroauric acid and
Chapter2
41
subsequentlyletthesolutionboilfor3.5min.TheNPsizewasintherangeof5.3±1.1nmforpureD2O
and9.0±1.2nm forpureH2O.The reduction in sizebyusingdeuteriumoxidewasattributed to the
increasedreducingstrengthofthecitrate(inD2O).
Uppal et al. synthesised gold NPs from 11 to 15 nm with trisodium citrate added to boiling
tetrachloroauric acid drop wise over a period of 6 min.98 The solutions were removed from the
temperaturebathusedtoheatthesolutionwhenthecolourofthesamplewentlilacandthesamples
wereobservedoveranumberofdays.Itwasfoundthatthepeakwavelengthdecreasedorblue-shifted
from540nmto522-523nmregardlessofwhetherthesolutionwasstoredindarkorlightconditions.
Uppal et al. also synthesised goldNPswith a number of different initiationmethods.99 This involved
initiating the reaction between tetrachlorauric acid and trisodium citrate using thermal heating,
sonolysis,microwaves andUV-light. These initiationmethodswere testedwith premixed solutions of
tetrachlorauricacidand trisodiumcitrate,where the trisodiumcitrateconcentrationwas increased to
varythereducingagent/stabilisingagentconcentration.11.0to11.9nmNPswereformedusingthermal
initiationwith little variance in sizewith increasing citrate concentration,11-17nmNPswere formed
usingmicrowaveswithadecreaseinsizewithincreasingcitrateconcentration,8nmNPswereformed
usingUV-lightinitiationwithanincreaseintheellipticalshapeoftheNPswithincreasingcitrateand17
to 18 nm NPs were formed using sonolysis and size changed minimally with increasing citrate
concentration.
Ojea-Jiménezetal.comparedthestandardmethodtoan inversemethod,which involvesadding
tetrachloroauric acid to hot sodium citrate.100 They found that theNPs synthesised using the inverse
methodwere smaller than those in the standard synthesiswhileNPs synthesized using the standard
methodatasodiumcitratetotetrachloroauricacidratioof13.6were36.6nmwhiletheinversemethod
gaveNPs9nminsize, loweringthesodiumcitratetotetrachloroauricacidratioto6.8gaveasmaller
disparity in size between the methods: 17.8 nm and 14.9 nm for the standard and inverse method
respectively.Byusingamixtureofsodiumcitrateanddicarboxyacetone(DCA)(90:10)theyfoundthat
thesurfaceplasmonresonancebandofthesolutiondevelopedquicker(from11minwithnoDCA,to5
minwithDCA)while therewasareduction from21.2to17.7nm insizewhenDCAwasaddedtothe
citratebefore addition to thehot tetrachloroauric acid. Byusing a 50:50mixtureofDCAand sodium
citrate,itwasfoundthattheresultantNPswerelargerandmorepolydisperse(25.9nm)andusingDCA
only resulted in no NPs being formed. Therewas a possibility of changes in size when changing the
sodiumcitrateDCAratiobecauseofchangingpH,andpHisaparameterknowntoaffectsizeofNPs.95
TocheckwhetherDCAhadarealeffectontheNPsize,thepHwasmodifiedusingsodiumhydroxide.A
decreaseinsizewasfoundwhenthepHwasadjustedtosimilarvaluestothatofexperimentswhichdo
notuseDCA,andtheresultantNPswerefoundtobesmallersuggestingthatDCAdoesindeedplayan
importantrole inthesynthesisofgoldNPs.Ojea-JiménezandCampanerasynthesisedgoldNPswitha
modificationofpHbyaddinghydrochloricacid to the tetrachloroauricacidbeforeadditionof sodium
citrate to themixtureat90°C,with thepHof6.5,5.6and4.7being tested.101TheNP sizedecreased
from36.6nmto16.1nmwithdecreasingpHvalue.
LiteratureReview
42
Sivaramanetal.synthesisedgoldNPsusingtheinversemethodbyaddingtetrachloroauricacidto
boilingtrisodiumcitratesolution.102Theyfoundthatatcitratetogoldratiosof<5,themeansizeofthe
NPwasnotsignificantlyaffected(52nmforinversemethod,56nmforTurkevichmethodataratioof1
to1)whereasforratios>5,therewasasignificantreductioninthesizeoftheNPs(10and16.1nmata
5.2ratio,and7.2and13.6nmataratioof20.8)whenusingtheinversemethod.Theyinvestigatedthe
possibilityoftheoxidationofcitratetoDCAbyboilingcitratebeforeadditiontotetrachlorauricacidand
comparing it to the standardmethod (citrate is notboiledbefore addition toboiling tetrachloroauric
acid)andtheinversemethod.TheyfoundthattheinversemethodyieldedsmallerNPsasexpected,but
thefirsttwomethodsyieldedsimilarsizessuggestingthatorderofreactantadditionisthecausefora
changeinNPsize,ratherthantheoxidationofcitratetoDCA.
Schulz et al. synthesized gold NPs of ~ 3.5 nm with a low dispersity of 5-6% using an inverse
method.103 Themethod was optimised by the use of sodium citrate/citric acid buffer to control pH.
EDTA was used to remove potential trace contaminations of metals such as zinc or copper, it was
suspectedthatthesecausedtheformationofgoldtriangularplatessincethemetalionscanaffectthe
growth of the Au NPs. It was found that the addition of EDTA reduced the formation of triangular
particlestopracticallyzero.
Wuithschicketal.synthesizedgoldNPsusingtheTurkevichmethodandinvestigatedtheeffectof
temperature,pHandvariationofreactantconcentrationsandorderofaddition.104Itwasobservedthat
there was a minimum size achieved of 12.2 nm when the synthesis was carried out 60°C and the
synthesistimedecreasedwithincreasingtemperature.ByvaryingtheinitialpHofthetetrachloroauric
acid,apHwindowbetween2.8and4wasfoundtoproducethebestdispersityofaround10%withthe
minimumsizeachievedof10.8nm.Thecitrateconcentrationsupposedlyhadlittleeffectaslongasthe
concentrationwasenough tomaintainaneutralpHaftermixing,whereas increasing tetrachloroauric
acid concentrations led to a decrease in the sizeof theNPs. Changing theorderof reactant addition
showed that adding tetrachloroauric acid to a larger volume of citrate gave smaller NPs than the
standardmethod.
Piella et al. synthesised sub-10 nm gold NPs based on a modification on the inverse method,
involving the use of tannic acid together with sodium citrate as a reducing agent.105 Highly
monodispersed3.6nmseedsweresynthesised(505nmpeakwavelength)andweresubsequentlyused
infurthergrowthstepstoobtainarangeofsizeinthesub-10nmrange.Parametricstudiesonoptimal
tannicacidtogoldmolarratio,temperatureandpHwereconducted,andtheoptimalvalueswerefound
tobe0.01,70°CandpH8 respectively forobtaining the smallestNPs. Sodiumcitrateand tannicacid
werealsotestedinisolationwhere10.9±1.3nm(518nmpeakwavelength)and4-12nmNPs(527nm
peakwavelength)wereobtainedrespectively,suggestingthereisasynergisticeffectbetweenthetannic
acid and sodium citrate, although the resultant 4-12 nm NPs from using tannic acid alone may be
becauseofinsufficientstabilisation.
Kettemannetal.synthesisedgoldNPsusingthestandardmethodandamodificationto improve
reproducibilityofthesynthesis.106WiththeimprovedsynthesistheywereabletosynthesiseNPsinthe
11-18 nm with less than 0.6 nm deviation in all cases. Similar to Schulz et al.,103 in the improved
Chapter2
43
synthesistheyusedabufferofsodiumcitrateandcitricacidtoachieveapHbetween5.5and6,which
was found to be the area inwhich the synthesis ismost reproducible because of the stability of the
HCit2- species. This species was suggested as the form of the citrate which is responsible for the
reductionofthegoldprecursor.
2.3.2 Microfluidicsynthesisofgoldnanoparticles
Wengetal.synthesizedgoldNPsusingaTurkevichmethodinanovelmicrofluidicdevice(Figure2-15).40
HexagonalNPs of around 35 nmwere obtained at a temperature of 115°Cwith reaction time in the
orderof2 to5min.Theyobserveda coalescenceof smallerNPs to form largerNPsbyanalyzing the
sample at a reaction of 2min. Thismechanismwas possibly due to the relatively quickmixing time
obtained in the reactor (reaction volumes were ca. 30-40 μl) which may have led to the quick
hydroxylationoftheAuCl4- speciestoAuClxOH4-x
- species,whichisknowntoreducethereactivityofthe
goldprecursor,possiblyleadingtoalowernucleationrateandincreasedgrowthrate,particularlywhen
thecitratetogoldmolarratiosarebetweenca.4-9.Thequickermixingtime(relativetoabatchprocess)
wouldhavereducedthetimeforreactionwhiletheAuCl4- speciesexistsinsolution.
Figure2-15:MicrofluidicdeviceusedtosynthesizehexagonalgoldNPs.40ReprintedfromRef.40withpermission
fromTheInstituteofPhysics.
Ftounietal.synthesizedgoldNPswiththeTurkevichmethodusingafusedsilicacapillaryandat-mixer
tomixchloroauricacidtosodiumcitratepriortointroductionintotheheatedcapillary(Figure2-16).107
In this case neither reagent was preheated before being mixed, and the residence times within the
capillarywerevariedbetween35and94s. Inthisresidencetimerange,theyobtainedNPsinthesize
rangebetween1.5and3nm,with largerNPsbeingobtainedat longer residence times. They further
investigatedtheeffectofsodiumcitratetogoldprecursormolarratio,andfoundaminimumataround
3.15.Theexplanationgivenisthatatlowercitrateratios,thereisanexcessofgoldprecursorwhichis
suggestedtoleadtoanincreasedgrowthofNPswhereasathighercitrateratiosthesizeissuggestedto
increase because of an increased ionic strength in solution. No discussion on the balance between
nucleation, growthand stabilization in termsofpHand reactivityof the respective reagents ismade,
LiteratureReview
44
althoughthisislikelytoplayasignificantroleinthefinalsizeoftheNPs.Temperatureeffectwasalso
investigated,andthesizeof theNPsdecreasedwith increasingtemperature in therangebetween60
and100°C.
Figure2-16:SynthesisofgoldNPsusingsyringepumps,t-mixerandfusedsilicacapillarytubingsetinahotplate.
Reproduced(Adapted)fromRef.107withpermissionofTheRoyalSocietyofChemistry.
Suganoetal.synthesizedgoldNPsbetween10-45nmusingay-typemicromixerwithapulsedflowof
chloroauricacidandsodiumcitratetomixthereagentswithinasmallchannelatroomtemperature.108
They investigatedhowmixingaffected the synthesisbyaltering thepulsing ratebetween50,100and
200Hz.Analysisofthemixingshowedthat increasingthepulsingrate improvedthemixingefficiency.
Theyusedtwocollectionmethods,inthefirstmethodthefluidwasallowedtocollectattheoutletof
themicromixerfor5minbeforecollectionandinthesecondmethodthefluidwascollectedforatotal
of 5 min through a 0.5 mm tube which was connected to the outlet of the micromixer. The
polydispersity of the resultant NPs showed the following behavior, collection method 2 produced
relativelylowerpolydispersitythancollectionmethod1forallpulsingfrequencies.Thisisattributedto
thereasoningthatcollectionmethod1allowsfluidthathastravelledthroughthemixerinitiallytomix
andreactwithfluidthatistravellingandexitingthemixeraftertheinitialperiod,resultingininteraction
ofthefluidwithdifferentconcentrationsovera5minperiod.Collectionmethod2sawareduction in
polydispersity from 50 Hz to 100 Hz, but further increase in frequency showed no improvements
suggestingthatfastermixingreducespolydispersityonlyupuntil100Hzandfurtherreductioninmixing
timehasnoeffect.ThesizeoftheNPswasfoundtoincreasewithahigherpulsingfrequency,suggesting
fastermixingresultedinlargerNPs.Thisismostlikelyduetothereactionrateofprecursordecreasing
withmixingastheAuCl4- speciesbecomesmorehydroxylatedoncontactwithcitrate.
Köhleret al. used a split and recombine type reactor (Figure 2-17) tomix tetrachloroauric acid,
ascorbicacid,PVA,sodiummetasilicateandFe(II)sulfatetosynthesisegoldNPs.109Becausetheyused4
pumpswhichwereindividuallycontrolledtofeedreactantstothereactorwhichhad4inlets,theywere
abletopreciselycontroltheconcentrationssimplybyalteringtheflowrates.Theyobtainedawidearray
of goldNPs using thismethod including small 5 nmgoldNPs,NPs arranged in a circle like structure,
aggregatesandlargerhexagonalaggregates.
Chapter2
45
Figure 2-17:Split and recombinemixer employed for the synthesis of goldNPs. Reproduced fromRef. 109with
permissionofTheRoyalSocietyofChemistry.
Wagner and Köhler synthesized gold NPs in a split and recombine reactor (similar to that shown in
Figure 2-17, but with 2 inlets instead of 4 using tetrachlorauric acid, ascorbic acid and
polyvinylpyrrolidone.Withincreasingflowrateinthereactor,adecreaseinNPsizewasobservedfrom
35nmataflowrateof1ml/minto24nmataflowrateof8ml/min.Alargersizeatlowerflowrates
wasattributed toapreferentialnucleationat thewallsof the reactor,whereadarkviolet filmwhich
gradually deepened and became opaque was observed. Supposedly this effect is lessened with
increasingflowratebecauseofanincreasedsheeratthewallsofthereactor,thoughifanequalsample
volumewastakenforeachexperiment, thismayhavebeenbecausethereactorwasoperatedovera
shorterperiodoftimeatthehigherflowrateandhencetherewaslesstimeforthedarkvioletfilmto
buildup.AnotherreasonforthesmallerNPsizeathigherflowrateswassuggestedtobeanincreased
nucleation rate because of more efficient mixing, which would lead to a quicker contact of excess
reducingagentwith theprecursor ions.Meandiameteranddispersityalsodecreasedwith increasing
ascorbicacidconcentrations,duetotheincreasednucleationrateobtainedatahigherreducingagent
concentration.SmallerandlessdisperseNPswereobtainedathigherpH(21nmatpH2.8,8nmatpH
9.5), attributed to the increasing redox potential of ascorbic acid at higher pH resulting in increased
nucleationrate.IncreasingPVPconcentrationalsoresultedinsmallerNPs,attributedtothesuppression
of growth around newly formed nuclei which occurs because of the large PVPmolecule slowing the
approachofgoldprecursor to thesurfaceofanuclei stabilizedbyPVP.Reductionof foulingwasalso
attainedusingtwodifferentapproaches,elevatedpHandsilanizationofthereactorwallstoproducea
hydrophobic surface. Both approaches achieved similar results in obtaining much higher peak
absorbancethanthestandardapproachatlowerpH.Köhleretal.andWagneretal.alsousedthesplit
andrecombinereactorstosynthesizegoldNPsusingborohydride.71,72
Shalom et al. synthesized small gold NPs using the Brust-Schiffrin method within a micromixer
whichsplitstwostreamsinto16substreamsandsubsequentlyinterdigitatesthestreamsforimproved
mixing(Figure2-18).110Tetrachloroauricacidwasreducedviasodiumborohydridewith1-dodecanethiol
being used as a stabilizing agent. NPs between 3-4 nm were synthesized within the reactor and a
significantreductioninsizeanddispersitywasachievedascomparedtosimilarconditionsusingabatch
reactor.Thiswasproposedtobebecauseofmoreefficientmixing,andsupposedlytherateofsodium
LiteratureReview
46
borohydrideadditionaffectstheNPsize.NPswillbemorepolydispersewhenadditionisslowbecause
nucleigeneratedearlyinthereactionareexposedtoanincreasedthiolconcentrationandthosenuclei
generated later as more sodium borohydride is added grow to a larger concentration because of
decreased thiol concentration. However, it is not mentioned that borohydride has an abundance of
electronsavailabletoreducethegoldprecursor,andhenceifmixingisslowthismayleadtoareduction
of many gold ions to metal via a single borohydride molecule, which subsequently doesn’t provide
enough surface charge to stabilize the formed nuclei leading to aggregation because of poorly
distributed sodium borohydride. Therefore one could argue that the polydispersity observed in the
batchisduetopoorlydispersedborohydride,leadingtolargerNPsproducedinitiallyratherthansmaller
NPs because of an increased thiol concentration. However, this would still explain why dispersity is
reduced in themicrofluidicsystem,sincethere isabetterdispersionofsodiumborohydrideearlier in
themixingprocess.
Figure2-18: InterdigitatingmicromixerusedbyShalometal. for the synthesisofgoldNPsvia theBrust-Schiffrin
method.110ReprintedfromRef.110,Copyright(2007),withpermissionfromElsevier.
Cabezaetal.synthesizedgoldNPswithinamicrochannelinwhichsegmentedflowwasgeneratedwith
anaqueousphase thatwas segmentedwitheitherair, tolueneor siliconeoil (Figure2-19).111Sodium
borohydridewasmixedwith tetrachloroauric acid and tetradecyltrimethylammonium bromide in the
aqueous phase and then segmented by the inert fluid at room temperature before being introduced
intoaheatedchannelwitha temperatureof100°C.They found increasing residence time resulted in
largerandmorepolydisperseNPs,withsizeincreasingfrom3.8±0.3to4.9±3.0nmataresidencetime
from 10 to 40 s, for a water-toluene systemwhere toluenewas the dispersed phase andwater the
continuousphase.Smallerslugsoffluidwereobtainedatalowerresidencetime,indicatingthatthese
types are more beneficial for controlling the size and polydispersity of the resultant NPs in the
continuousphase.This isprobablybecauseofthedecreasedsizeofthecontinuousphaseslugswhich
would result in faster mixing and therefore decreased polydispersity. Considering this, they further
Chapter2
47
investigated the effect of changing the dispersed phase fluid on NPs synthesized in the continuous
aqueousphaseandfoundthatairasthedispersedphaseproducedthesmallestandmostmonodisperse
NPs(2.8±0.2nm)followedbytoluene(3.8±0.3nm)andsiliconeoilproducedabimodaldistribution
(7.8 ± 6.5 nm and 15.5 ± 3.1 nm). This was determined to be because of how well the continuous
aqueous slug phase recirculatedmaterial,with air producing the highest circulation rate followed by
tolueneandsiliconeoil.Thiswasdownthehydrodynamicswithinthemicrochannelandthedifference
inmaterialpropertiesoftheinertdispersedphase.
Figure2-19:Segmented flow inmicrochannelemployed for synthesisofgoldNPsbyCabezaetal.Reprintedwith
permissionfromRef.111.Copyright(2012)AmericanChemicalSociety.
Tsunoyama et al. synthesized small gold NPs of around 1 nm by reducing tetrachlorauric acid with
sodium borohydride in the presence of polyvinylpyrrolidone in an interdigitating micromixer (Figure
2-20).112Smallerclusterswereobtainedwhentheflowratewasincreasedandwhentheconcentration
oftheprecursorandreducingagentwasincreased.
Figure2-20:InterdigitatingmixerusedbyTsunoyamaetal.forthesynthesisofgoldNPs.Reprinted(adapted)with
permissionfromRef.112.Copyright(2008)AmericanChemicalSociety.
LiteratureReview
48
ThegoldNPsynthesisinmicrofluidicdevicesstudiespresentedshowthewiderangeofpossibilitiesand
potential benefits that microfluidic devices offer for the synthesis of NPs, and also highlight the
importanceofmasstransferincontrollingsizeanddispersityoftheNPs.
2.3.3 Mechanisticstudiesofgoldnanoparticlesynthesis
Oneof themorecommonlyheldviewsof themechanismforgoldNP formation isaburstnucleation
process,asdescribedbyLaMerandDinegar,113inwhichthenucleiareformedinashortperiodofhigh
nucleationrate,causingthesupersaturationtodropbelowtherequiredlevelfornucleationandintoa
growthphasewhichresultsinmonodispersedparticles.Turkevichetal.conductedastudytoinvestigate
themechanisms involved in thenucleationandgrowthofgoldNPs in thechloroauricacidandcitrate
system.114Thenucleationmechanismwassuspectedtohavefourdistinctregions:inductionperiod,an
auto-acceleratingperiod,alinearperiodandadecayperiod.Briefly,theinductionperiodwasbelieved
to be caused by the oxidation of citrate ions into acetonedicarboxylate ions which when reaching a
sufficientlevel,inducesnucleationthroughthereductionofprecursor.Theauto-acceleratingportionis
suspected tobe causedby the increase inproductionof acetonedicarboxylate ionspast the speedat
whichtheycanbeutilisedforreduction.Thelinearproportioniswhennucleationachievesaconstant
rate and the limiting factor is the speedatwhich theprecursor is reduced to formnuclei. Thedecay
period occurs when nucleation gives way to the growth of the NPs, i.e. the precursor begins to be
utilisedatthesurfaceofexistingnucleisurfacesratherthannucleatingnewparticles.Itisimportantto
note thatat thedecaystage, itwasobserved that less than5%of thegoldprecursorandcitratehad
beenconsumed,sothedecayperioddoesn’tseemtobeduetotheexhaustionofreagents.
Biggsetal.investigatedthecolloidalstabilityofthegoldNPsandtheroleoftheAuCl4- ionandthe
citrateion.115UsingAFMmeasurements,itwasshownthatthecitrateionbindsontothesurfaceofgold
causingarepulsiveforce,butwhentheAuCl4- ionwasintroduced;therepulsiveforceisgreatlyreduced
suggestingthatthecitrateisdisplacedbyAuCl4- .ThepreferentialattachmentofAuCl4
- ontothesurfaces
of the gold NPs causes a reduction in the repulsive force of the NPs leading to flocculation. This
flocculation is thecauseof thepurplish-bluecolorusuallyobserved in theTurkevichsynthesis,andas
theAuCl4- isconsumedthroughsurfacereduction,thereisasubsequentincreaseinrepulsionforcesat
thesurfaceasmorecitrateattachesleadingtotheformationofawell-dispersedgoldsol.Injectionofa
mixtureof goldprecursorand citrateontoa goldplate surface thathadbeenpreviously equilibrated
withcitrateshowedthatthesurfacepotentialwentfromrepulsivetoattractiveandremainedattractive
becauseoftheslowreductionoftheAu(III)ontothesurfaceofthegoldplate.116Furtherworkwasdone
on this by Wall et al. using the same techniques but with a thiol coated cantilevers to prevent NP
formation, and the results revealed that the behaviorwas consistentwith amultilayer adsorption of
citrateontothesurfaceofthegoldandbecauseofthelowchargedensity,therewasmostlikelysome
counter-ionpairingwiththecitrate.117ThepossibilityoftheAuCl4- ionadsorbingwithacounter-ionor
withaCl-displacementwaslikelyaccordingtotheresultsofthestudy.
Chapter2
49
Chow and Zukoski suggest a growth mechanism were large fluffy NPs are formed in the initial
stages of the reaction and then subsequently these larger agglomerated NPs break down to form
smallerandmoremonodispersedNPs.91TheysuggestthattheconcentrationofsmallNPsincreasesand
larger NPs decreases through the course of the reaction. The proposed mechanism can be briefly
described as follows. Aurate ions are reduced by citrate to form gold atoms which form clusters
consisting of gold metal, with adsorbed citrate and aurate ions. Since aurate ions are preferentially
bindedovercitratetothesurfaceofthegold,thereisalowersurfacepotentialwhichresultsinaloose
aggregation.Over time the aurate ions are reducedonto the surface to grow theNP, resulting in an
increase in surfacepotential eventually leading to anelectrostatic repulsionwhich causes the loosely
aggregatedNPstobreakoff intosmallerNPs.ThisdescribeshowthelargefluffyNPseventuallybreak
downtoformamonodispersedandstablesol.
Pei et al. monitored the evolution of gold at various reaction times by monitoring the
concentration of the precursor species (AuCl4- ) and the total gold concentration, which allowed the
concentrationofmetallicgoldtobeinferred.93 ItwasfoundthatthereductioninAuCl4- ionwasfaster
withahighercitratetoprecursorratiowhich isnaturalsincethecitratebehavesasareducingagent.
NanowiresremainedinsolutioniftherewasAuCl4- ionsstillinsolutionandinthecaseofratios0.2and
0.3, nanowires persisted until at least 180 min, supporting the idea that the AuCl4- does lead to an
aggregationofNPsassuggestedbythestudiesabove.AssoonasAuCl4- wasconsumed,thebluecolorof
thenanowiresdisappearedandtheredcolorofsphericalNPswasobtained.HenceexcessiveAuCl4- ion
was important innanowire formation,and itwasobservedthatexposuretohydrogengascausedthe
breakdownofnanowires intoNPspresumablybecauseofa reductionofexcessAuCl4- . It is suggested
thatthemechanismisasfollows:nucleationofsmallgoldparticles,subsequentattachmentoffurther
AuCl4- ontotheparticlesurface,thisAuCl4
- ionresults inanattractiveforcebetweenNPswhichcauses
themtosticktogether,followedbygrowthofnanowiresandlarge2-Dstructuresofnanowires.
Kimlingetal.foundthatthecrystalsizeswithincitratepreparedgoldNPwereusuallysmallerthan
theparticlediameterindicatingapolycrystallinenature.94Thissupportstheideaofnucleationfollowed
by growth through coalescence. They propose a four step process described here briefly as: cluster
formation,agglomerationofclustersintoNPswithadefinedsurfaceplasmonresonance(SPR),fusionof
these NPs into larger entities and finally there is a significant amount of atoms in solution which
continuetoaggregateleadingtotheir incorporationintothelargerentitiesresultinginahomogenous
density and improved symmetry of NPs resulting in a blue shift of the SPR. This growthmechanism
appliedforUV-assistedNPformationandusingascorbicacidasareducingagent,andwasbasedonthe
movementof theSPRwhichstarts froman initialphasewhichblueshifts, followedbyaredshiftand
then finally a blue shift. This behavior could also be explained by a loose aggregation of NPs due to
unreduced AuCl4- on the surface , explaining the red shift, and eventual surface reduction leading to
uniformstableNPs,explainingthesubsequentblueshift.
LiteratureReview
50
Kumaret al. proposed amodel for the formation of goldNPs based on reactions they deduced
from evidence in the literature pertaining to stoichiometry of reactions and suggestions for the
oxidationofcitrateintotheDCAmolecule.118Briefly,themechanismisdescribedbyareductionofthe
Au3+ ion toAu+via theoxidationof thecitratemolecule toDCA,whichsubsequently formsacomplex
withtheAu+.TheDCAorganizestheAu+anddisproportionationofthegoldionsintogoldatomsoccurs.
Theseatomspickup furtherAu+ ionsto form largeraggregatesandfurtherdisproportionationoccurs
leading to larger gold clusters and subsequent formation of gold nuclei. Based on this, a model is
presentedwithmassbalancesforthespecies involved,whichaccounts forthemassofauricchloride,
aurouschloride,citrate,DCAandacetone(acetoneisalsoproposedasareagentwhichcanreduceauric
chloride to aurous chloride). A population balance model is also included so that the number
concentrationofparticles,whichplaysaroleintheconsumptionofprecursorthroughsurfacegrowth,
can be deduced. Kinetic rate constants for each step of the proposedmechanism are then obtained
throughfittingtotheexperimentaldatapresentedbyFrens.89Areasonablefittotheexperimentaldata
in the literature is obtained, suggesting that the proposedmechanism is able to account for at least
some of the physical mechanisms occurring in the synthesis of gold NPs in the Turkevich synthesis.
Interestinglythereisnosuggestionofparticleaggregationtoformwirelikestructuresandsubsequent
reduction intomonodisperseNPswhich is observedexperimentally bymany studiesdiscussed in this
section.
Jietal.proposeamechanismfortwodifferentsetsofratiosofcitratetogoldprecursor,onefor
lowcitratetogoldratios(<3.5)andtheotherforhighcitrateratios(>3.5),showninFigure2-21.95At
high citrate ratios,where the pH is closer to neutral there is nucleation followed by a slowdiffusion
controlledgrowth,evidencedbysmallsphericalNPsatthe initialstagefollowedbygrowth into larger
spherical NPs. At the lower citrate ratios, dense nanowires with fatter heads and thinner tails were
observedwhicheventuallyleadtosphericalNPs.Itissuggestedthisisfarmorelikelybecauseofintra-
particleripeningratherthanthroughthecleavingofmaterial fromthenanowireswhichthengrowto
largersphericalNPswhichissuggestedbyPongetal.119Thisisbecausetheevolutionfromnanowiresto
spherical NPs takes much longer than the reduction of gold precursor into gold metal, hence there
wouldbea lackofmonomersavailable forgrowthofanycleavedNPsandmostof theprecursorhas
alreadybeenutilizedtoformthenanowirestructures.Thereisalsoanobservedfatteningofthehead
andthinningofthetailovertimewhichsupportstheideaofintra-particleripening.Themechanismwas
found to be controlled by the reactivity of the precursorwhichwas determined by its speciation. By
tuningthepH, itwasfoundthatthereactivityofprecursorcorrespondedwiththeswitchingpointsof
thegoldioncomplexesi.e.reactivityfellasthepHincreasedasaresultofincreasingthehydroxylation
of thegold ion fromAuCl4- toAuCl3OH
- toAuCl2OH2-.This reactivity is suggested tobe the reason for
nanowirelikestructuresatlowerpHandsphericalNPsathigherpH.Sivaramanetal.alsofoundtherate
ofreductionofgoldspecieswashighestwhenthetetrachloroauricacidwaspresentinitsmostreactive
form(i.e.AuCl4- ).102Theextentofprotonation/deprotonationofcitratemaybeimportantalso,sincethe
pointatwhichaminimuminsizeoccurredwasatpHaround6.5,correspondingtothethirdpKavalueof
6.4forcitricacidwhereitiscompletelydeprotonated(Ct3-).
Chapter2
51
Figure 2-21:Mechanism suggested by Ji et al. for gold NP synthesis. Two pathways exist dependent on pH.95
ReprintedwithpermissionfromRef.84.Copyright(2007)AmericanChemicalSociety.
Polteetal.observedthenucleationandgrowthofthegoldNPsusingtheTurkevichsynthesisat75and
85°CthroughtheuseofUV-Vis,SAXSandXANESanalysis.120Incontrasttomechanismswhichsuggest
theformationof largerNPswhicheventuallyformsmallerandmoresphericalandmonodisperseNPs,
therewerenolargerNPsobservedinthisstudy.Insteadthemechanismisproposedtobeconsistingof
fourmainsteps: the reduction followedbynucleationofgold, thecoalescenceofgoldnuclei to form
largergoldNPs,slowgrowthof the largergoldNPsthroughreductionofprecursorontoexistingNPs,
and finally amore rapid growth onto existingNPs possibly because of an autocatalytic effect (Figure
2-22). Polteet al. alsoobserved a coalescenceof smallerNPs to form largerNPswhenusing sodium
borohydrideasareducingagent.121GoiaandMatijevicusediso-ascorbicacidtoreducetetrachloroauric
acid and found a coalescence of gold NPs of a crystal size of 30 nm which formed much larger
micrometer sized particles.122 Park et al. modelled the process of burst nucleation followed by
aggregationofprimaryparticlestoformlargersecondaryparticlesandthemodelwasfoundtosensitive
tothesurfacetensionoftheparticlesandsinglet-singletaggregationrate.123
Figure2-22:MechanismofgoldNPformationsuggestedbyPolteetal.Reprinted(adapted)withpermissionfrom
Ref.120.Copyright(2010)AmericanChemicalSociety.
Wuithschicket al. investigated themechanism from the speciation of reagents aspect and answered
somequestionspertainingtothecontrolofthesizeoftheNPs.104Theypresentasuggestedmechanism
forgoldNPformationfortheTurkevichsynthesis,basedonthatproposedbyPolte(Figure2-23).124The
main suggestion is that the speciation of the AuCl4- controls the nucleation of the gold NPs which
LiteratureReview
52
ultimatelydeterminestheirfinalsize.ThedegreeofhydroxylationoftheAuCl4- affectsthereactivityas
highlighted by Jiet al.,95 and that the Turkevichmethod producesmonodisperseNPs because of the
change inspeciationoftheprecursorwhichcreatesaseparationofnucleationandgrowthasAuCl4- is
bothconsumedandtransitions intomorehydroxylated formsaspHgoes fromacidic toneutral.They
notethatoncethenucleation isover, thekineticsofgrowthhas littleeffectonthe finalsizewhich is
determinedinthenucleationphase,suggestedtobeoftheorderof30slongfortheconditionstested.
Theyemphasizethat,contrarytothemajorityofstudieselucidatingthemechanismofgoldNPsynthesis
usingtheTurkevichmethod,thereisnoaggregateoragglomerateformationleadingtotheformationof
wire like structures according to the SAXS analysis by Polte et al.120 The bluish-gray color which is
observed in the Turkevich synthesis is suggested to be due to an optical alteration of gold NPswith
precursor ions adsorbedonto the surface rather thanbecauseof an aggregationor agglomerationof
NPs. Plech et al. synthesised NPs using a modified Turkevich method in which X-rays were used to
inducenucleation.125They introducedamixtureofchloroauricacidandtrisodiumcitrate intoanX-ray
capillary, where it was exposed to X-rays and formed NPs. In the early stages of the process, they
suggestabimodaldistributioncentredaround7.8and3.9nmat38s,and9and4.6nmat83s.Following
this period, the data suggests that the particles are close in distance and that there is clustering of
particles to bigger aggregates (20-40 particles). Finally after 2h, the data can be fitted by isolated
particleswithafewclustersstillpresent.ThefinalsizeoftheNPswasnotdifferenttotheaggregates
detectedintheearlierstages,supportingtheideaofcoalescenceofsmallerNPstoformlargerones.
Figure2-23:MechanismofgoldNPformationintheTurkevichsynthesissuggestedbyPolte.Ref.124–Publishedby
TheRoyalSocietyofChemistry.
Kettemannetal. investigated the importanceof the speciationof thecitrate species in theTurkevich
synthesis.106 Depending on the pH, the protonation/deprotonation of the citrate species changes,
increasingpHfromacidictobasicdeprotonatesthespecies.TheydeducedthattheHCit2- wasthelikely
species which reduced the AuCl4- and if the synthesis is manipulated to ensure a reproducible
concentrationofHCit2- ,thesynthesisbecomeshighlyreproducible.HAuCl4
- wasalsopronetodifferences
Chapter2
53
from batch to batch, it was found that boiling and refluxing the precursor for 1 h enabled
reproducibility.
Muchwork has been carried out to understand themechanisms involved in goldNP formation,
with particular focus on the reduction of gold precursor using citrate at elevated temperatures. The
colorchangesobserved in thesynthesiscanbest summarize themechanisms. Initially there isacolor
change from the yellowofAuCl4- to clear,which ismost likely the transitionof theAu3+ ion to some
intermediatesuchasAu+assuggestedbyKumaretal.118 Followingthis,apurple-bluecolor isusually
observed.Theconsensusisthatthiscolorarisesfromflocculationorlooseaggregationofnewlyformed
goldnuclei,observedtobelargefluffyNPs,91orwirelikestructures.93,95Theexplanationofthisseems
to arise from thepreferential binding of gold precursor to the surface of gold nuclei over the citrate
molecule,ashighlightedbytheworkofBiggsetal.andWalletal.115-117Thisthenleadstothecolorof
thesolutiondeepeningandeventuallyturningrubyred,indicatingtheformationofmonodispersedgold
NPsof10-20nm.This is suggested tobecausedby theslowsurface reductionof thegoldprecursor,
whicheventually leads to increasing stabilityofNPsas citratebegins to attachonto the surface. This
increasedstabilitycausesarepulsion,whichenablestheflocculatedNPstobreakapart.Thereseemsto
beanagreementthatacoalescenceofnuclei inthe initialphasetoformsmallgoldNPs, indicatedby
their polycrystalline nature and the SAXS studies by Polte et al. and Plech et al.120, 125 There is a
suggestionthattheformationofwirelikestructuresorlargerNPsdoesn’texistinsolutionandthatthe
purplebluecoloriscausedbyanopticaleffectofgoldprecursorbindingontothesurfaceofgoldNPs.104
However, it is still seems like flocculation isa strongpossibilitybecauseof thepreferentialbindingof
goldprecursoronthesurfaceofgoldNPs,whichwouldcreateattractiverather thanrepulsive forces.
ThespeciationofthecomponentsintheTurkevichsynthesishavealsobeenidentifiedtobeimportant
indeterminingthemechanisms,reactionsratesandresultantNPsize.
54
3 Synthesisofsilvernanoparticlesinamicrofluidiccoaxialflowreactor
The disadvantage of using reactorswith a high surface area to volume ratio is that there is a higher
chanceoffoulingandaccumulationofmaterialonthewallsofthechannel.Hence,microreactorswhich
are not easily susceptible to accumulation ofmaterial on theirwalls are desirable. There are various
studieswhichimplementeddifferentstrategiestoavoidfouling,suchasthesegmented-flowapproach
and the reduction of wall interactions through pH alteration and surface silanization of the channel
walls.76, 126, 127One typeof design that alleviates theoccurrenceof fouling is the coaxial flow reactor
(CFR)whichallowsaninnerstreamofreagenttobesurroundedbyanouterstream,creatingareaction
interfacebetweenthestreams.35,36,128-130TheCFRislesssusceptibletofoulingbecausethereductionof
precursorionsandnucleationofmetalNPsoccurattheinterfacebetweentheinnerandouterstreams,
rather thannear the channelwalls. This reduction in fouling relieson the flowprofileof the channel
remaininginastablelaminarregime.Thishasadverseeffectswithregardstomixingefficiency,unlessa
reductionofdiffusiondistancesorsufficientcontacttimeisapplied.
ExamplesofCFRsarefoundinthestudiesbyTakagietal.whoproducedtitaniaNPsusinganaxial
dual-pipe microdevice.128 Using various flow conditions and tuning of solvent properties such as
viscosityandsurface tension, the flowcharacteristics in theirdevicewere tunedand led to increased
controlovermass transfer.Themass transferof the reagents in themicrodevicedetermined the size
and monodispersity of the NPs which was demonstrated by using different solvent types, tube
diametersandprecursorconcentrations.Makietal.employedaconcentric-axledual-pipemicroreactor
to synthesize zirconia NPs,128, 129 and showed that increasing residence time resulted in larger NPs.
Abou-Hassanetal.usedacoaxialflowdeviceoperatedunderlaminarflowconditionstosynthesizeiron
oxide, goethite and Fe2O3@SiO2 core/shell NPs.35, 36, 130 7 nm iron oxideNPswere synthesized in the
coaxialflowdeviceandpolydispersitydecreasedfrom35%inatypicalbatchsynthesistoca.20%inthe
coaxial flow device.130 Goethite NPs were synthesized using the coaxial flow device as a nucleation
sectionbeforealongresidenceloopinatemperaturecontrolledbath.ThegoethiteNPswereproduced
witharesidencetimeofaround15minasopposedtoseveralhoursusingtraditionalbatchmethods,
andwere also smaller than thosemade in batch vessels.35 Similarly, amultistep approach using the
coaxialflowdeviceswasusedtosynthesizeFe2O3@SiO2core/shellNPswithinamuchshortertimescale
ascomparedtobatch.36Limetal.operatedaCFRintheturbulentregimetoproducevariousNPssuch
asironoxidewithanabilitytotunethesizethroughchangingtheflowconditions.131Schüleinetal.used
amicrocoaxial-injectionflowmixerforthesynthesisofnickelNPs.Theprocessforsynthesizingnickel
NPsquicklyblockedcommerciallyavailablemicrostructuredmixersandthecoaxial-injectionmixerwas
usedinapilotplantscaleprocessmainlybecauseitcouldavoidfoulingandblockageissues.132
Synthesis of silver NPs using silver nitrate, sodium borohydride (with sodium hydroxide) and
trisodiumcitrateinacoaxialflowreactorinthisstudyhasnotbeenreportedintheliterature.Shirtcliffe
etal.reportedthesynthesisofsilverNPsusingsodiumborohydrideandsodiumhydroxideasareducing
agent for silvernitrateusingaPTFE flowchamberandalsousingEppendorfpipettes tomix reagents
directlyintoplasticcuvettes.65Wagneretal.synthesizedsilverNPsinasplitandrecombinemixerusing
Chapter3
55
borohydrideasareducingagentforsilvernitrateanddespitefoulingwereabletocontrolthesizeofthe
NPsbytuningtheborohydrideratio,obtainingNPs intherangeof10-20nmforborohydridetosilver
nitrate ratios ranging from 3 to 40.71 Further studies using borohydride as a reducing agent for the
reductionofasilverprecursor to formsilverNPs in theaqueousphaseweremainly inbatchreactors
withNPsizesrangingbetween1-100nm.5,64,67,69
Themainobjectiveofthiswork istostudyhowthesizeandmonodispersityofsilverNPscanbe
controlled by changing the operating parameters of flow rate and concentration of ligand (trisodium
citrate) andprecursor (silver nitrate). Themotivation for using a CFRwas to avoid fouling issues and
blockages,anissuewhichwasencounteredwhensynthesisingNPsusingthesamechemicalsystemina
split and recombine type micromixer (see Appendix B, Figure B1 for images of fouling in such a
micromixer).
3.1 Experimental
3.1.1 Chemicals
Silver nitrate (AgNO3, 0.01 M stock solution), trisodium citrate (HOC(COONa)(CH2COONa)2 ·2H2O,
powderform)andsodiumborohydridesolution(NaBH4,~12wt%in14MNaOHstocksolution),were
obtainedfromSigmaAldrichCompanyLtd.,UK.Allchemicalswereusedwithoutfurtherpurificationand
solutions were prepared with ultrapure water (resistivity 15.0 MΩ·cm) and diluted to the desired
concentrationsbeforebeingintroducedintotheCFR.
3.1.2 ExperimentalSetup
Syringe pumps (Pump 11 Elite OEM Module, Harvard Apparatus) were used in the experiments to
deliverthetwostreamsofsilvernitrate/trisodiumcitratesolutionandsodiumborohydridesolutionto
theCFR.TheCFRusedinthisstudyconsistedofasmallinnertubewhichwasinsertedintothecenterof
alargeroutertube(Figure3-1).Boththeinnerandoutertubesweremadeofglassandinnertubesof
twodifferentinternalchanneldiametersof0.556mm(0.8mmexternaldiameter)and0.798mm(1.09
mm external diameter) were used in separate experiments. The outer tube had an internal channel
diameterof2mm.ThedistancefromtheinnertubeoutlettotheoutletoftheCFRwas130mm.A1.59
mmPTFEtubesleevewithasuitable innerdiameter isplacedaroundtheinnertubetokeepthetube
stablewithinthelargeroutertube.ThecapillarieswereconnectedtogetherusingaT-piececonnector
(0.508mmthru-hole,UpchurchScientific)whichwasdrilledtoasizeof2mminternaldiameter(I.D.)to
allow the inner tube to be inserted into the outer tube. The flow ratio between the inner andouter
stream,Qout:Qin,wasfixedat1:1andtheCFRoperatedinthelaminarflowregimeforallexperiments.
Synthesisofsilvernanoparticlesinamicrofluidiccoaxialflowreactor
56
Figure3-1:Schematicof thecoaxial flowreactorsetup. Insertshows flowvisualizationof laminar flow inside the
coaxialflowreactorwithbluedyeflowingthroughtheinnertubeandwaterflowingthroughtheoutertube.
3.1.3 Synthesisofsilvernanoparticles
Atypicalsynthesiswasas follows:silvernitrateandtrisodiumcitrateweremixedtogether insolution
andintroducedviatheinnertubeoftheCFR,whilethesodiumborohydridesolutionwasintroducedvia
theouter tubeof the reactor. Fresh trisodiumcitrate solutionwasmadebydissolving thepowder in
ultrapurewater before each synthesis. All reactionswere carried out at room temperature (typically
between 22-24 °C). All concentrations, unless otherwise stated, are concentrations of the individual
streams being introduced into the CFR. Since the sodium borohydride was stored in 14 M sodium
hydroxide, the concentration of sodium hydroxide was 3.21 times higher than the stated sodium
borohydrideconcentrationinallcases.
The literature shows that borohydride can supply up to eightmolar equivalents of electrons to
reduce the silver ions to silver metal.133-135 The amount of electrons released from the borohydride
molecule depends on whether the formed H+ ions react with H- ions from the molecule to form
hydrogengas,whichwouldlimittheelectronsavailableforreducingmetalions.Highlybasicconditions
reducetheformationofhydrogengashencefreeingupmoreelectronsforreductionofthemetalion.In
this study sodium borohydride was stored and used under basic conditions which should produce a
higher amount of electrons for reducing the metal ions. Samples were stored for months without
noticeableformationofhydrogenbubbleswhereaswhensodiumhydroxidewasnotusedtoarrestthe
hydrolysisreaction,hydrogenbubblesformedwithinthesamplesafterashorttime.Theactualamount
ofelectronsutilizedtoreducesilverisunknown,thoughitissuggestedbyShirtcliffeetal.thatsixofthe
availableeightelectronsareutilizedtoreducesilverionsunderbasicconditions.65
Theprimarypurposeofsodiumborohydrideistoreducethesilverionstosilvermetal;however,it
alsoaffects thegrowthof theNPs.Thesurfacechargeof theparticles isaffectedbyborohydrideand
hydroxide ion adsorption and it is well established that the conversion of borohydride to borate in
aqueoussolutionresultsinchangesintheparticlecharge.68,136Underbasicconditionstheconversionof
borohydrideslowsdown.137,138However,trisodiumcitrateisusedastheprimarystabilizingagentinthis
study.Theconcentrationsofthereducingagent(sodiumborohydride)andligand(trisodiumcitrate)in
Chapter3
57
relation to the precursor concentration (silver nitrate) determine the nucleation, growth and
stabilizationof theNPs.Sincetheconcentration is important indeterminingthesizeanddispersityof
resultant NPs, the mass transfer during the reaction becomes important as this dictates the
concentration at which the reaction takes place at any given time during the reaction. The
reproducibility of the synthesis in the CFR over a range of flow rateswas assessed by repeating the
synthesis 4 times for each flow rate (Figure B2 in Appendix B showsUV-Vis analysis of the repeated
syntheses).
3.1.4 Characterizationofnanoparticles
ThesilverNPswereanalyzedwithinanhourofthesynthesisusingaUV-Visspectrometer(USB2000+
Spectrometer andDT-Mini-2-GS light source, OceanOptics). Transmission electronmicroscope (TEM)
imageswerecapturedusinga JEOL1200EX iimicroscopewitha120kVaccelerationvoltage.Carbon
coatedcopperTEMgridswerepreparedwithinanhourof synthesisbypipettingapproximately15µl
sampleontothegridandallowingittodryatroomtemperature.Itwasimportanttokeepthedroplet
smallenoughtoonlypartiallycoverthegrid.Thishastwobeneficialeffects.Firstly,thedryingtimeof
grid is shorter because of a lower volume of solution. The second reason relates to the well-known
‘coffee ring’ effect, where particles are pushed towards the edges of a droplet that is evaporating,
meaningthattheseareaswillcontainaggregatedNPswhicharenotrepresentativeofthesamplebut
rather an artefact of sample preparation. When the grid is prepared in such a way, the ring of
aggregatedNPscanbephysically identifiedonthesmallcoppergridandeasilyvisiblewhenanalysing
thesampleusingTEM.Thisallows thedifferentiationbetweenaggregates thatare in the sampleand
those that are artefacts of sample preparation. The TEM images have in the inset particle size
distributions, average diameter, d, standard deviation, δd, indicating polydispersity and number of
particlescountedtoobtaintheparticlesizedistribution,n.PEBBLESsoftwarewasusedforthecounting
andsizingofTEMimagesofthesynthesisedNPs.
3.2 Investigationonvarianceinnanoparticlemorphologyusingparametricstudies
3.2.1 Effectoftotalflowrate
Experimentswerecarriedout intheCFRwithan innertubechanneldiameterof0.798mmatvarious
flowrates.TheCFRwasusedtosynthesizesilverNPsatconcentrationsof0.2mMsilvernitrate,0.2mM
trisodium citrate and 0.6 mM sodium borohydride at total flow rates ranging from 1 ml/min to 14
ml/min.Themolarratioofsodiumborohydridetosilvernitrateis3:1,buttheratioofelectronstosilver
ionsismanytimesgreaterthanthis.Figure3-2showsTEMimagesofsilverNPssynthesizedat1ml/min,
2.5ml/min,8ml/minand14ml/minalongwiththeirrespectiveparticlesizedistributions.Theparticle
sizesynthesizedatthe lowerflowratesof1ml/min(3.1±1.6nm)and2.5ml/min(3.3±2.2nm)are
statisticallysmallerthanthoseatthehigherflowratesof8ml/min(5.6±2.7nm)and14ml/min(5.4±
3.4nm)while thepolydispersity steadily increases from±1.6nm to ± 3.4nm.However, theparticle
sizesat8ml/minand14ml/minwerenotfoundtobestatisticallydifferentusingat-test(p=0.209).
According toMie theory, forNPs in the rangebetween6-10nm thepeakabsorbance increases
with size (foragivenconcentration)but thepeakwavelengthdoesnot shift significantly.139Fromthe
Synthesisofsilvernanoparticlesinamicrofluidiccoaxialflowreactor
58
UV-Visanalysis(seeAppendixB,FigureB3),thepeakwavelengthdoesnotchangesignificantly(between
392-393nmforallflowrates)inagreementwiththeory.However,thepeakabsorbanceofthespectra
decreasesminimally from1.35 to 1.27 (statistically different, p < 0.01)with increasing flowrate even
thoughTEMshowsanincreasingaveragesizeofNPsfrom3.1nmto5.4nm.Thiscanbeexplainedby
theincreasedpolydispersityobservedathigherflowrateswhereNPslargerthan10nmareformed.This
results in increasedfullwidthathalfmaximumoftheresonancepeaksathigherflowrates,andsince
themassofsilverusedineachexperimentisconstant,thepeakabsorbancedecreases.
Figure 3-2:TEM images and particle size distributions of silverNPs synthesized at different total flowrates. A: 1
ml/min,B:2.5ml/min,C:8ml/minandD:14ml/min.Concentrationofsilvernitrate0.2mM,trisodiumcitrate0.2
mM,sodiumborohydride0.6mM.0.798mminnertubeI.D.
The concentrations of precursor, reducing agent and ligand are important factors in controlling
nucleation and growth and hence size of the synthesized NPs. The evolution of the concentration
profiles along the length of the reactor is controlled bymass transfer, i.e. themixing characteristics.
Mixing in theCFRoccurs by diffusion, since it operates in the laminar flow regime for the flow rates
employed.Wecanuseasimplerelationship:140
Chapter3
59
t≈(x/2)2
2D
wheretistime,xisthethicknessofthestreamandDisthediffusioncoefficient,toestimatethetime
required for the reagent to diffuse significantly along the radial direction in the CFR (i.e. the mole
fractionbecomesapproximatelyuniformacrossthechanneldiameterintheabsenceofreaction).Using
adiffusioncoefficientof1.7x10-9m2/sforAgNO3,141andastreamthicknessof1mm(halfthediameter
oftheoutertube),atimeof73siscalculated.Usingadiffusioncoefficientof3.5x10-9m2/sforNaBH4
inNaOHsolution,142atimeof36siscalculated.Thevolumeofthereactoris0.4ml,yieldinganaverage
residencetimeof1.75-24.5sforflowratesbetween1-14ml/min.Hence,completemixingdoesnot
occur because the residence times are shorter than the characteristic diffusion times. At higher flow
rates,minimalmixing occurs in the channel as themolecules have very little time for diffusion. The
majority of the reagentmixing in this case is at the outlet of the CFR where droplets form and are
collected.Ahigh reactant concentration in the streams ismaintained in theCFRbecauseof the slow
mixing. This is confirmed when considering the Peclet number of the conditions tested. For a
characteristiclengthof2mm(innerdiameteroftheoutertube)thePecletnumbervariesbetweenca.
3000-42000 for sodiumborohydride and ca. 6000-87000 for silver nitrate. This indicates that for the
flowconditionstested,themasstransportisdominatedbyadvection.Thisisbecausetheresidencetime
withinthechannelisquiteshortcomparedtothetimeneededfordiffusionalongthecross-sectionof
thechannel.Evengiventheadvectiondominatedmasstransport(becausetheresidencetimeisquicker
thandiffusiontime),thesizeanddispersityoftheNPsisstillsignificantlyaffectedwithvaryingflowrate
whichhighlightsthespeedatwhichthereactionstakeplacewiththelimiteddiffusiontimeavailable.
Literature suggests that reaction and nucleation in silver NP synthesis have very fast kinetics
evidencedbythetimescalesofNPappearanceinstudiesbyPolteetal.whichsuggestsreductionoccurs
inlessthan200ms,68andTakesueetal. inwhichnucleationisobservedinlessthan1ms.82Thus,the
reactionislikelytooccuratorveryclosetotheinterfacebetweenthetwostreams,atleastintheinitial
partofthereactorandathighflowrates.Theinnerandouterstreamsbehaveasreservoirsofprecursor
and reducing agent respectively which supply the necessary reagents at the interface region for
reduction of silver ions to silver metal and subsequent clustering and stabilization. Comparing the
averageresidencetimeswithcharacteristictimesfordiffusion,wecanconcludethatathighresidence
time(24.5s)mostsilvernitrateintheinnerstreamwouldbeconsumed,becauseofthefastkineticsand
sufficientdiffusiontimeoftheborohydride(36s).Thisislikelynottobethecaseatlowresidencetime
(1.75s).
Higher flow rates resulted in the appearance of larger silver NPs. This may be rationalized as
follows.The reactionoccurringat the interfacewithin theCFRresults in silveratomsbeing formed in
thatregion.Theconcentrationofsilveratomsincreasesuntiltheyformclusters(synthesismechanisms
in the literature strongly suggest that the metal atoms form clusters which coalesce to form small
NPs).68, 82, 143 A smaller amount of silver precursor is consumed and smaller amounts of clusters are
formedwithintheCFRatlowresidencetime.Ahigheramountofsilvernitrateisleftunreactedatthe
outletoftheCFR,wheredropletsformandmixingandconsumptionofreagentscontinuewithinthem.
Synthesisofsilvernanoparticlesinamicrofluidiccoaxialflowreactor
60
Mixing there is through recirculation patterns as the droplets form. Even though these convective
patterns promote faster mass transfer than diffusion, they result in non-homogenous concentration
areas. Thus, the larger amount of remaining silver ions reacts and moves to areas where the
concentration of silver nitratemay be high relative to borohydride. The resultant NPs would have a
lowersurfacechargebecauseofthelowamountofborohydrideionswhichadsorbontotheirsurface.
ThisinturnallowstheNPstoaggregateandgrowtoalargersize,sinceitisestablishedthatadecrease
ofsurfacechargeoftheNPs(mostlikelybecauseofborohydrideiondepletion)causesNPstogrowtoa
larger size.136 The appearance of higher amount of larger NPs, further gives rise to the increased
polydispersity observed. The increasing Peclet number (with increasing flow rate) also quantifies the
diminishing significance of the diffusionmechanism on themass transportwith increasing flow rate,
furthersupportingtheideathatthereislessreactionoccurringwithinthechannelathigherflowrates.
Aninterestingobservationwasthatalayerofsilverformedontheinnerwallneartheoutletofthe
innertubethroughwhichthesilvernitratewasflowing.Thisfoulingoccurredontheouterwallofthe
innertubewhensilvernitratewasflowingthroughtheoutertube.Foulingonlyoccurredonthesideof
theinnertubewheresilvernitratewasflowingi.e.asilvernitraterichzone(seeFigureB4andFigureB5
in Appendix B for images of fouling on the channel walls). The reason fouling did not occur on the
sodiumborohydride sideduringexperimentsmaybebecause silvermetal atomsproduced in sodium
borohydriderichzoneswouldhaveahighnegativesurfacechargeandhencecannotgrowanddeposit
on the negatively charged glass channelwall in the sodium borohydride stream.144 This is consistent
withthelargerNPsobtainedathigherflowrates,whichincreasesthepolydispersity.Anotherpointto
noteisthatcitratemoleculeswerealsoavailableinthesilvernitratestream,buttheydidnotprevent
thedepositionofsilvermetalontheinnertube.
3.2.2 Effectoftrisodiumcitrateconcentration
TheeffectofsurfactantconcentrationontheNPmorphologywasinvestigatedbyvaryingthetrisodium
citrateconcentration.Thesilvernitrateandsodiumborohydrideconcentrationswerekeptconstantat
0.1mMand0.3mMrespectively.Theroleofsodiumborohydrideandwhy itwas insuchhighexcess
wasdiscussedpreviously.Trisodiumcitrateconcentrationwasvariedbetween0.025mMand1.5mM.
Duetothesehighcitrateconcentrations,thesilvernitrateandsodiumborohydrideconcentrationwas
reduced (though still using a 3 to 1 ratio of sodium borohydride to silver nitrate) to keep the ionic
strengthaslowaspossible.Theflowrateofeachstreamwas1.25ml/minmakingatotalflowrateof2.5
ml/min. Figure 3-3 shows TEM images and particle size distribution of the silver NPs synthesized at
variouscitrateconcentrations (0.025mMto0.75mM)withaveragediameterandstandarddeviation
foreachconcentration.ItcanbeseenthatthepolydispersityandaveragesizeoftheNPsdecreaseswith
increasingcitrateconcentrationbecauseofthereductionintheamountoflargefusedNPspresent.
Figure3-4showstheUV-VisspectraofthesilverNPsamplesforeachcitrateconcentrationtested.
The resonance peaks at lower citrate concentration are characterized by a lower plasmon peak
absorbance and a higher absorbance at longer wavelengths; and as the citrate concentration is
increasedupto0.25mMtheplasmonpeak increasesandthepeakathigherwavelengths is reduced.
The tail in the resonancepeak is indicativeof thepresenceof larger fusedNPs,asconfirmedbyTEM
Chapter3
61
imaging at lower citrate concentrations, suggesting citrate plays a key role in reducing formation of
largerfusedNPs.Increasingtheconcentrationofcitrateabove0.5mMupto1.5mMdoesnotalterthe
synthesized silver NP characteristics. This shows that a minimum citrate concentration is needed to
obtain relativelymonodispersedNPs and to prevent the formation of larger fusedNPswith irregular
shapes.Henglein andGeirsig also showed in a studyusing radiolysis for the reductionofAgClO4 that
largerfusedNPswereformedwheninsufficientamountsofcitratewereused(theyalsoshowedlarge
fused NPs when the citrate concentration was too high).83 Further qualitative analysis of the TEM
images shows that at lower concentration, there are smaller NPs which surround the larger NPs.
BecauseofthehighcontrastofthelargerfusedNPs,thesesmallerNPSmaybehardertoidentifysince
theydonotyieldashighofa contrast.Table3-1 shows the fullwidthathalfmaximum (FWHM)and
peakwavelengthofthesilverNPssynthesisedatthevariouscitrateconcentrations,obtainedfromthe
UV-VISspectra.AsexpectedtheFWHMdecreaseswithincreasingconcentration(lowerFWHMindicates
lowerdispersity),inlinewiththeTEMimagesshowingdecreasedpolydispersitywithincreasingcitrate
concentration.Interestingly,thepeakwavelengthshowsincreasewithincreasingcitrateconcentration
(higherpeakwavelengths indicate increasingNP size). This is obviouslynot the caseaccording to the
TEMimagesaswellasthelargeshoulderpeaksintheUV-Vis,butitdoessupporttheideathatatlower
citrate concentrations, there is a possibility of smaller NPs existing (in particular, smaller than the
smallestNPspresentinthesynthesisathighercitrateconcentration).
Synthesisofsilvernanoparticlesinamicrofluidiccoaxialflowreactor
62
Figure 3-3: TEM images and particle size distributions of silver NPs synthesized at different trisodium citrate
concentrations. A: 0.05mM , B: 0.1mM , C: 0.25mM andD: 1.5mM. Concentration of silver nitrate 0.1mM,
sodiumborohydride0.3mM.Totalflowrate2.5ml/min,0.556mminnertubeI.D.
Chapter3
63
Figure 3-4:Absorbance peaks of silver NPs synthesized at various trisodium citrate concentrations in the range
0.025-1.5mM.Total flow rate2.5ml/min, concentrationof silvernitrate0.1mM, sodiumborohydride0.3mM.
0.556mminnertubeI.D.
Table3-1:FWHMandpeakwavelengthofsilverNPsatvarioustrisodiumcitrateconcentrationsintherange0.025-
1.5mM.Totalflowrate2.5ml/min,concentrationofsilvernitrate0.1mM,sodiumborohydride0.3mM.0.556mm
innertubeI.D.
Citrateconcentration(mM) FWHM(nm) Peakwavelength(nm)
0.025 96 381
0.05 91 379
0.1 84 385
0.25 77 389
0.5 77 393
1 75 392
1.5 77 392
3.2.3 Effectofsilvernitrateconcentration
TheeffectofprecursorconcentrationontheNPsizeandpolydispersitywasinvestigatedbyvaryingthe
concentrationofsilvernitrate.Thetrisodiumcitrateandsodiumborohydrideconcentrationswerekept
constantat0.5mMand0.3mMrespectively.Thesilvernitrateconcentrationwasvariedbetween0.05
mMand0.4mM.Sodiumborohydrideisexpectedtoreduceallsilverionsevenatthehighestprecursor
concentrationusedsinceitisexpectedtosupplyanexcessofelectronsforreduction.Theflowrateof
eachstreamwas1.25ml/minmakingatotalflowrateof2.5ml/min.TheBeer-Lambertlawappliedfor
Synthesisofsilvernanoparticlesinamicrofluidiccoaxialflowreactor
64
synthesizedAgNPshavingpeakabsorbance valuesbelow1.5. Therefore, sampleswerediluted if the
absorbancewasabovethisvaluefornormalization.
Figure 3-5 shows TEM images of the silver NPs along with their respective particle size
distributions.Thesizeincreasedfrom3.7±0.8nmto9.3±3nmwithincreasingsilverconcentration.It
can be seen that small and relatively monodisperse NPs form at the lowest concentration with a
transitionintoamorepolydispersebimodaldistributionoccurringat0.15mMsilvernitrate.Thehigher
polydispersitymayberelatedtoalargerextentofreductiontakingplaceinthedropletsattheexitof
the CFR, as discussed earlier. Figure 3-6 shows the UV-Vis spectra for each synthesis carried out at
various silver nitrate concentrations. The peak absorbance increases with increasing concentration
showingalineardependence.TheincreaseismostlikelybecauseofahigherconcentrationofNPsand
increasedaverage sizeof theNPs (absorbance increaseswith size forNPsbelow10nm).139Table3-2
shows the FWHM and peak absorbance of the silver NPs synthesised at various silver nitrate
concentrations,obtainedfromtheUV-Visspectra.Atthelowestconcentrationthespectraisdiffuseand
hasalargerpeakwavelength(whichindicateslargeandpolydisperseNPs).However,theNPSare3.7nm
accordingtoTEM.Atthissize,theNPswouldshowverylittleabsorbance,sotheabsorbanceshownin
thespectracouldbe fromsmallamountsof largerNPswhichwouldhaveamuchhigherabsorbance.
The largeFWHMcouldbebecause the largerNPs thatexist in thesolutionarepolydisperse.Thiscan
explain to an extent the FWHM and peak wavelength at a concentration of 0.15 mM also. With
increasing concentration after 0.15 mM, the FWHM and peak wavelength shows little variation
indicatingthatthereisminimalchangeinNPsizeandpolydispersity.
Asdiscussedpreviously,highamountsofsilvernitrateinthedropletsformingattheoutletofthe
CFRcanleadtosilvernitraterichzones(duetoimperfectmixing)whichleadtolargerNPsforming(as
wellasincreasedpolydispersity).Inthecaseofthelowestsilvernitrateconcentration(0.05mM)there
isalowamountofsilvernitrateinthedropletsattheoutletoftheCFR,leadingtoalowerpossibilityof
silver nitrate rich zones. At higher silver nitrate concentrations there is a larger amount within the
droplets at theoutlet of theCFR, increasing the likelihoodof reactionsoccurring in silvernitrate rich
zonesleadingtoanincreasedformationoflargerNPsandpolydispersity.
Chapter3
65
Figure 3-5: TEM images and particle size distributions of silver NPs synthesized at different silver nitrate
concentrations.A:0.05mM,B:0.15mM,C:0.25mMandD:0.4mM.Concentrationoftrisodiumcitrate0.5mM,
sodiumborohydride0.3mM.Totalflowrate2.5ml/min,0.556mminnertubeI.D.
Synthesisofsilvernanoparticlesinamicrofluidiccoaxialflowreactor
66
Figure3-6:AbsorbancepeaksofsilverNPsynthesizedatvarioussilvernitrateconcentrationsintherange0.05-0.4
mM.Total flowrate2.5ml/min,concentrationsof trisodiumcitrate0.5mM,sodiumborohydride0.3mM.0.556
mminnertubeI.D.Inset:Silvernitrateconcentrationvs.peakabsorbanceforresonancepeaks.
Table3-2:FWHMandpeakwavelengthofsilverNPsatvarioussilvernitrateconcentrationsintherange0.05–0.4
mM.Total flowrate2.5ml/min,concentrationsof trisodiumcitrate0.5mM,sodiumborohydride0.3mM.0.556
mminnertubeI.D.
Silver concentration (mM) FWHM(nm) PeakWavelength(nm)0.05 88 402 0.15 76 399 0.2 67 397
0.25 64 395 0.3 68 396 0.4 66 395
3.3 Conclusions
TheperformanceofaCoaxialFlowReactor(CFR)wasinvestigatedforsilverNPsynthesis.Thereaction
wasconfinedtotheinterfacebetweentheinnerandouterstreamsresultinginthegenerationofsilver
clusterswhich increase in concentration along the length of theCFR.Hence the residence time is an
important parameterwhichdetermines the amount of the precursor consumedbefore the advective
mixingwithinthedropletsattheendoftheCFR.Residencetimedeterminestheamountofsilvernitrate
consumedintheCFR,anditisshownthatshorterresidencetimeswherelesssilvernitrateisconsumed,
results in larger NPs being formed. This is possibly because reactions in silver nitrate rich zones (i.e.
reactor outlet at high flow rates) lead to the formation of NPs with low surface charge which have
reducedstabilityandhencehaveatendencytogrowbycoalescence.Forthesamereason, increasing
Chapter3
67
silvernitrateconcentration,whilemaintainingaconstantresidencetimeresulted inan increaseofNP
size and polydispersity. Increasing citrate concentration reduced the amount of larger fused NPs
resulting in a smaller average size and reduced polydispersity. Such fused NPs disappeared above a
citrate concentration of 0.25 mM. Increasing the concentration of citrate past this amount did not
decrease the size of the NPs further, indicating that a threshold is reached. The CFR offers a simple
method tocontrol thesizeof theNPsbecauseof itsuniquedesign.Bykeepingadvectivemixing toa
minimum,thereactionismaintainedattheinterfaceregionbetweenthetwostreams.Thisavoidsthe
problem of blockage of the channels by containing the reaction away from the channelwalls, unlike
other types of microfluidic devices where the reaction can take place close to the channel walls.
68
4 Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
The synthesisof silverNPsusing sodiumborohydrideasa reducingagent typically leads to smallNPs
becauseborohydrideisastrongreducingagent,howeverittendstogivepolydisperseNPs.Ourprevious
studyusingacoaxialflowreactorproducedsilverNPsinthe4-10nmrangeandshowedthatmixingis
animportantparameterwhenusingborohydrideasareducingagent.145Splitandrecombinetypeflow
reactors71 and PTFE flow chambers65 have been utilized to synthesize silver NPs using borohydride
obtainingNPs in the rangeof 10-20nm for borohydride to silver nitrate ratios ranging from3 to 40.
BatchreactorsynthesesofsilverNPsusingborohydrideasareducingagentproducedNPsinthe1-100
nmrange.5,64,67,69
Anon-confinedimpingingjetreactor(IJR)withcharacteristicdimensionsinthemicrometerrange
isemployedinthisstudytoinvestigatehowmanipulatingmasstransferaffectstheNPcharacteristics,in
particularthesizeanddispersity.IJRisaparticularlyusefuldevicetouseinthesynthesisofNPssinceit
doesn’t clog because of absence of channel walls. Fouling is particularly problematic in microfluidic
devices since their small dimensions make them potentially vulnerable to blockage, and various
strategiesintheliteraturehavebeenusedtoavoidfoulingsuchassegmentedflow,126coaxialflow128,145
andsurfacesilanization.127
The IJR has foundmost common application in bipropellant liquid rocket engines146 but it also
presents some very distinct advantages for reactions which produce solids. Mahajan and Kirwan
characterizedthemixingtimesinanIJRusingtheBournereactionscheme.147Theysubsequentlycarried
out a precipitation reaction for the synthesis of lovastatin crystals and compared the size of crystals
obtainedwhen themixing timewas i) slowerand ii) faster than the induction time forcrystallization.
They concluded that amoremonodisperse particle size distribution (PSD)was obtainedwhenmixing
timewasfaster.Theyalsoconcludedthat ifthemixingtimeisfasterthantheinductiontime(elapsed
time period between achievement of supersaturation and the appearance of crystals),148 further
improvements to themixing time do not significantly affect the resultant crystals. Erni and Elabbadi
studiedthehydrodynamicpropertiesoftheIJRandcharacterizedvariousflowregimesbeforeusingthe
IJR for an acid induced crystallization of sodium benzoate. They found a decrease in both size and
dispersity with an increasing jet velocity.149 Kumar et al. employed an IJR for the synthesis of
nanocrystallineMgO,aprocesswhichyieldsarigidgelwhichcanquicklyclogmicrochannels.Achange
inimpingementangleandvelocityofthejetswasfoundtoaffectthesurfaceareaofthenanocrystalline
MgO formedand foreachparameter therewasanoptimumthatexisted toobtainmaximumsurface
area.Itissuggestedthatthisoptimumexistsbecauseofachangeinhydrodynamicsintheimpingement
zone,wheretoohighaflowrateorimpingementangleledtonon-optimalmixingconditionsresultingin
areducedcontacttimeforreaction.150Hosnietal.usedimpingingjetreactorsforthesynthesisofZnO
NPs,comparingresultsagainst those formed inabatchstirredvesselandaT-mixer typereactor.The
sizeoftheNPssynthesizedineachreactordependedontheenergydissipationusedformixing,withthe
Chapter4
69
IJRhavingintermediateenergydissipationversustheothertwotypesofreactorsandhenceproducing
anintermediatesizeofZnONPs.151
Confinedimpingingjetreactors(CIJR)areasimilartypeofreactorfortheuseinreactionsinvolving
precipitationofsolids.JohnsonandPrud’hommeemployedaCIJRfortheprecipitationofhydrophobic
organic activesNPs; importantly theCIJRallows forhomogeneous conditionsprior toprecipitationof
NPs (i.e. formation of block copolymer NPs and nucleation and growth of organic actives) which
ultimately separates the effect ofmixing on the final size of the NPs.152 The NP size decreasedwith
increasingmixing efficiency up until a ‘breakpoint’was reached, atwhich point the size of the block
copolymerNPs remained constant. They also observed the same effectwhen using a different block
copolymer,andthe‘breakpoint’correspondedtoaDahmköhlernumberof1.153Marchisioetal.studied
theprecipitationofbariumsulfateNPsinaCIJRandsimilarlyobservedareductioninparticlesizewith
increasingmixingintensity,butnofurtherimprovementswereobservedwhenthecharacteristicmixing
timescalereachedthereactiontimescale.154ReductioninsizeofprecipitatedbariumsulfateNPswith
increasingmixing intensitywasalsoobservedbySchwarzerandPeukertusingaT-mixer.155Siddiquiet
al.usedaCIJRforthesynthesisofironoxideNPswheretheyfoundareductionfrom800nmto200nm
in agglomerate size by increasing the flow rate under high concentration conditions.156 These studies
showthe importanceofmixingforreactionswhich leadtoformationofNPs,sincetheytypicallyhave
reactiontimesthataresmallandhenceneedverygoodmixingefficiencytoapproachmonodispersity.
According to the literatureabove, a reduction inNP size is reportedwith increasingmixingefficiency
attributed to increasing supersaturation levelswhichwould favor a higher nucleation rate andhence
smallerparticles.
The IJRused in thisstudypresentsaconvenientplatformfor thesynthesisofsilverNPsbecause
the reactor has no channel walls in the reaction zone, eliminating the need for lengthy cleaning
procedureswhich are needed in traditionalmicrochannels because of fouling. The IJR can synthesize
silverNPs inahighly repeatablemanner,associatedwith the lackof foulingwhichcanoccur inother
typesofreactorswheresurfacesarepresentnearthereactionzone.Themainobjectiveofthisworkis
toinvestigatetheeffectofmixingonthesizeanddispersityofsilverNPssynthesizedusingtwodifferent
ligands: i)trisodiumcitrateandii)polyvinylalcohol(PVA).Thehydrodynamicsandmixingefficiencyof
theIJRhavealsobeenstudied.
4.1 Experimental
4.1.1 Chemicals
Silver nitrate (AgNO3, 0.01 M stock solution), trisodium citrate (HOC(COONa)(CH2COONa)2·2H2O,
powderform,PVA(molecularweight89000-98000Da),sodiumborohydridesolution(NaBH4,~12wt%
in 14 M NaOH stock solution), potassium iodide (KI, 99%) potassium iodate (KIO3, 98%), boric acid
(H3BO3,99.5%)andsulfuricacid (H2SO4,3Mstocksolution)were obtainedfromSigma.Allchemicals
were usedwithout further purification and solutions were preparedwith ultrapure water (resistivity
15.0MΩ·cm).
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
70
4.1.2 ExperimentalSetup
Syringepumps(Pump11EliteOEMmodule,Harvard)wereusedintheflowexperimentstodeliverthe
reagents to the IJR. Figure 4-1 shows a schematic representation of the IJR, which consisted of two
stainlesssteeltubeswithequal internaldiametersof0.25mmor0.5mmalignedatanangle,q=48°.
This was achieved by inserting the tubes into a perspex plate which had 1.5mm (depth andwidth)
channelsmilledata48°angleusingaRolandEGX-400engravingmachine.Theouterdiametersofthe
tubingwas1.59mm.Spacingbetweenthecentersofthefluidjetsatthepointoftubeoutlets,d,was3
mm.
Figure4-1:Representationoftheimpingingjetreactor.
4.1.3 NanoparticleSynthesis
Amixturewasmadebycombiningsilvernitrateandligand(trisodiumcitrateorPVA)intoasolutionand
diluting to the appropriate concentration for input 1 (Figure 4-1) and diluting sodium
borohydride/sodiumhydroxidesolutiontotheappropriateconcentrationforinput2.Sincethesodium
borohydridewas stored in 14M sodiumhydroxide, the concentration of sodiumhydroxidewas 3.21
times higher than the stated sodium borohydride concentration in all cases. The syringe pumps
deliveredreagentstoinput1andinput2attheappropriateflowratestoproducethedesiredjets.As
the two jets collide, the reagents mix followed by subsequent reaction and formation of silver NPs.
Reactionswerecarriedoutat roomtemperature (22-24°C).All concentrationsstatedare thoseat the
inletsbeforeanymixingofreagentsoccurs.
4.1.4 CharacterisationofNanoparticles
ThesilverNPswereanalyzedusingaUV-Visspectrometer(USB2000+SpectrometerandDT-Mini-2-GS
lightsource,OceanOptics)between45to60minaftersynthesis(thesignalofthesampleswerestable
in thiswindowof time). SilverNP sampleswere dilutedwith additional ultrapurewater to bring the
absorbanceintoasuitablerange(i.e.obeyingtheBeer-Lambertlawandavoidingsaturationofthelight
detector)andthedatawerenormalizedsothatthemaximumabsorbanceintheflowraterangetested
representedanabsorbanceof1.TransmissionelectronmicroscopeimageswerecapturedusingaJEOL
1200 EX ii microscope with a 120 kV acceleration voltage. Carbon coated copper TEM grids were
preparedbetween45to60minaftersynthesisbypipettinga5µlsampleontothegridandallowingit
todryat roomtemperature.Carewas takentonotcover theentireTEMgridwithsamplebut rather
only a portion of it (as discussed in Chapter 3). Particle size distributions (insets for each TEM image
Chapter4
71
presented)havethefollowingnomenclature:disaveragediameter,δdisthestandarddeviationofthe
NPdistributionandnisthenumberofparticlescountedtoobtaintheparticlesizedistribution.PEBBLES
softwarewasusedforthecountingandsizingofTEMimagesofthesynthesisedNPs.
4.2 Mixingcharacterizationmethods
For single phase mixing, two types of experimental technique can be considered for mixing
characterisation:dilution-basedmethodsandreaction-basedmethods.157
4.2.1 Dilution-basedmethods
Thesemethodsinvolvethecontactingofaclearliquidwithacoloureddyeandobservingthechangein
concentration of the coloured dye along the channel. Mixing is considered complete when the
concentration of the coloured dye is uniform across the diameter of the channel. There are two
limitations to this kind of characterisation. The channel must be optically transparent for visual
observations using amicroscope to bemade. Because the observation ismade perpendicular to the
channel, theconcentration that isvisuallyobserved isactuallyanaverageof theconcentrationof the
dyeoverthedepthofthechannel.Examplesofthistypeofcharacterisationare:Mixingconcentrated
dye with colourless water solution,158 and dilution of fluorescent species such as fluorescein and
rhodamineB.50
4.2.2 Reaction-basedmethods
Threetypesofreactionexistinthisbranchofcharacterisation:
1. Acid-basereactions:
2. Reactionsproducingacolouredspeciesproduct
3. Competingparallelreactions
Thefirsttwomethodsalsorelyonopticalobservationoftheproduct. Acid-basereactionsrelyonthe
presenceofapHindicatorsuchasmethylorange,bromothymolblueandphenolphthalein. Reactions
producing coloured species include the reductionofpotassiumpermanganate in alkalineethanol and
theironrhodanidereaction.Boththesemethodsgivesimilardisadvantagestodilution-basedmethods
(notbeingabletoobtainconcentrationsacrossthedepthofthechannel,butobtainanaverageofthe
concentrationoverthedepthofthechannel).
Competingparallelreactionsrelyonparallelreactionswhichcompeteforasinglereactant.Oneof
thesereactionswillbefasterthantheother.Thedegreetowhichtheslowerreactionoccursdepends
on the measure of mixing. There are two types of these reactions used for measuring the mixing
performance in microreactors: the Villermaux-Dushman reaction and the second Bourne reaction
scheme (the simultaneous diazo coupling of 1- and 2-napthol with diazotized sulfanilic acid). Both
methods involve measuring the concentration of specific products using UV-Vis spectroscopy, which
allowscertainparameterstoquantifymixingtobecalculatedsuchassegregationindex.Forthisstudy,
theVillermaux-Dushmanreactionswerechosentocharacterizemixing.
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
72
4.3 Villermaux-Dushmanreactionscheme
TheVillermaux-Dushmantestreactionsystemisawell-knownmethodtocharacterizemixingefficiency
inareactor.159Mixingtimewithinamicromixingdevicecanbeobtainedusingthisreactionsystemby
tuningtheconcentrationsappropriately.TheVillermaux-Dushmannreactionsystemcanbedescribedas
follows:Astoichiometricdefectofsulphuricacidisintroducedtoamixtureofborate,iodideandiodate
ions.Thefollowingreactionsoccur:
H2BO3-+H+↔H3BO3 (4.1) instantaneous
5I-+IO3-+6H+↔3I2+3H2O (4.2) veryfast
I2+I-↔I3- (4.3)
Reaction1ismuchfasterthanreaction2,andsincethereisastoichiometricdefectofacid,onlyreaction
1occurs if there isperfectlyefficientmixing. Inpractice, reaction2 increasesas themixingefficiency
decreasesbecauseiodideandiodatewillreactwithacid ifthelocalconcentrationoforthoborateions
hasbeendepleted(thisonlyoccursifthemixingisnotperfectlyefficient).Reaction2producesiodine,
whichfurtherreactswithiodideionstoformtriiodide(reaction3).Adefectofacidamountisneeded
becausethemeanpHmustbewithinanarrowrangeclosetothevaluefortheiodinedismutationpH
(pH*).IfthepHislowerthanpH*,iodinecanforminacidsolutionirrespectiveofmicromixingeffects,
while on the other hand the pH cannot be too basic since formation of iodine becomes
thermodynamicallyunstable.159Bymeasuringthetriiodideconcentration,theextenttowhichreaction2
occurs can be measured, hence allowing the mixing efficiency to be estimated. Triiodide exhibits a
strongpeakat353nmenablingitsconcentrationtobecalculated.Theextinctioncoefficientoftriiodide
was found to be 23209M-1·cm-1 at 353 nm,which is similar to the values reported in the literature
(detailsonthisprocedurecanbe found inAppendixAandcalibrationcurve isshown inFigureA2).159
The micromixing quality is quantified by the segregation index which is discussed in detail in other
studies.159, 160 The segregation index can be linked to the mixing efficiency through the use of a
phenomenologicalmixingmodel.
4.4 Theinteractionbyexchangewiththemeanmixingmodel
Theinteractionbyexchangewiththemean(IEM)modelwasusedtoestimatethemixingtimefromthe
experimentalresults,andcanbeexpressedasfollows:161,162
dCk,1dt
= <Ck>-Ck,1tμ
+Rk,1 (4.4)
dCk,2dt
= <Ck>-Ck,2tμ
+Rk,2 (4.5)
<Ck>=αCk,1+ 1-α Ck,2 (4.6)
Chapter4
73
whereCisconcentrationofspeciesk,<Ck>istheaverageconcentrationofthespecieskacrossacidand
bufferstreams,Risthereactionrateofspeciesk,tμisthemixingtimeandαisthevolumefractionof
theacidstream.Equations3.4and3.5representtheacidstream(orinjection)andthebufferstream(or
volumeinthetank)respectively,andhencethesubscriptfollowingkintheseequationsrepresenteither
the acid (1) or buffer (2) stream. These equations form thedifferentialmass balance equations that
need tobesolved toobtaina theoretical concentrationof triiodide (or segregation index) foragiven
mixingtimeaccordingtothemodel.Thesegregationindex,𝑋2,isdefinedas:
Xs=YYST (4.7)
whereYistheratioofacidmolenumberconsumedbyreaction2dividedbythetotalacidmolenumber
injected:
Y=2 I2 + I3
-
α(Ho+) (4.8)
andYSTisthevalueofYwhenthemixingprocessisinfinitelyslow(perfectsegregation):
YST=
6(IO3- )0
(H2BO3- )0
6(IO3
- )0
(H2BO3- )0
+1
(4.9)
A curve of mixing time vs. segregation index can be obtained by setting a mixing time and the
concentrationofreagents,followedbysolvingforarangeofmixingtimes.Thedifferentialmassbalance
equationswere solvedusinggPromsModelBuilder3.7.1.Anexample code fora flow reactor canbe
seen inAppendix A. By solving the IEMmodel for a range ofmixing times, segregation index can be
plottedagainstmixingtime.Byperformingtheexperimentsusingtheappropriatereactor,thetriiodide
concentration for the given set of conditions can be obtained and the segregation index can be
determined. The corresponding mixing time is then determined by using the plot generated by the
model i.e. seewhichmixing time correspondswith the segregation index determined experimentally
fromtheplot.Therelationshipbetweenmixingtimeandsegregation indexfor theconcentrationsets
usedtocharacterisemixingtimeintheIJRscanbeseeninFigure4-2.
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
74
Figure4-2:Segregationindexvs.mixingtimeestimatedusingtheIEMmodelfortheVillermaux-Dushmannreaction
schemefor0.5mm(blueline)and0.25mm(redline)IJRs.ConcentrationsofbuffersolutionwereKI:0.032M,KIO3:
0.006M,NaOH:0.045M,H3BO3:0.09MinbothIJRs.ConcentrationofH2SO4was:0.017Mand0.02Mfor0.5mm
and0.25mmIJRrespectively.
4.5 Experimentaldeterminationofmixingtimeintheimpingingjetreactors
TheVillermaux-Dushmannreactionsystemwascarriedoutinthe0.5mmand0.25mmIJRs.Forinput1,
0.017Mand0.02Mofsulphuricacidwereused(correspondingto0.034Mand0.04MH+)forthe0.5
mmand0.25mmIJRrespectively.Thedissociationofsulphuricacidisconsideredtobealmostcomplete
forpHabove4,whereasifthereactionstakeplaceatalowerpHthentheassumptionofcompleteacid
dissociationwouldresultinoverestimatingthemixingefficiency.CommengeandFalknotedthatthepH
of the solution (acid and buffer combined) will increase as mixing takes place resulting in more
dissociation, while suggesting that if the reaction takes place at pH 2.5, the mixing time will be
underestimated by 25%.160 Input 2, which is considered the buffer solution, had a concentration of
0.032Mpotassiumiodide,0.006Mpotassiumiodate,0.045Msodiumhydroxideand0.09Morthoboric
acid.Figure4-3showsmixingtimevs.Webernumberobtainedforthe0.5mmand0.25mmIJRs.
Chapter4
75
Figure 4-3:Mixing time calculated by theVillermaux-Dushmann reaction scheme vs.Weber number for 0.5mm
(blacksquares)and0.25mm(redtriangles)IJRs.Concentrationofacidsolution(input1)was0.017Mand0.02M
H2SO4 for 0.5mmand 0.25mm I.D. tubing IJR respectively. Concentrations of buffer solution (input 2)were KI:
0.032M,KIO3:0.006M,NaOH:0.045M,H3BO3:0.09MinbothIJRs.
TheWebernumberisdefinedas:163
We=ρu2dσ
whereρ isthedensityoffluid,uisaveragevelocityoffluidbeforeimpact,disthediameterofthejet
(takenasthetubeinternaldiameter)andσisthesurfacetensionofthefluid.Valuesoffluidproperties
usedare those forwaterat roomtemperature.TheWebernumber represents the ratioof inertial to
surface tension forces, and is useful in describing what regime the IJR is operating in. The Weber
number range for the 0.5 mm and 0.25 mm IJRs was 13-65 and 32-176 respectively. The Reynolds
numberisdefinedas:
Re=ρudμ
whereμisthedynamicviscosity.TheReynoldsnumberrepresentstheratioofinertialtoviscousforces,
anddescribeswhether the flow isa laminar, intermediateor turbulent regime.TheReynoldsnumber
rangeforthe0.5mmand0.25mmIJRswas680-1520and760-1770respectively.FromFigure4-3itcan
beseenthatthemixingtimeforbothreactorsissimilarwhenoperatingundersimilarWebernumbers.
ThisindicatesthattheWebernumberisakeyparameterindeterminingthemixingefficiencyofanIJR.
Erni and Elabaddi propose that the mixing time for an impinging jet is proportional to the Weber
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
76
numberaccordingtoapowerlawrelationship(tm∝We-0.75).149FittingthedatainourstudyforWe<
80, the following relationshipwas found, tm∝We-0.89.Mixing time decreaseswith increasingWeber
numberuptoavalueofca.90whereitreachesaminimumvalueofaround1ms.IncreasingtheWeber
numberpastca.90usingthe0.25mmIJRresults inagradual increase inmixingtime. Ingeneral, the
specificpowerdissipationwithinareactorcanbeusedtoestimatethemixingtimeandseveralstudies
have showna correlationbetween the two.52, 160, 161 The specificpowerdissipation for the IJR canbe
estimatedusingthefollowingrelationships:147
𝜖 = 456 (4.10)
𝑃 = 89(𝑚8𝑢89 + 𝑚9𝑢99) (4.11)
𝑉 = @AB
Cℎ (4.12)
where 𝜖 is overall energy dissipation per unit mass [W/kg], P is rate of overall mechanical energy
dissipation [W] assumed to be equal to the kinetic energies of the two impinging jets,𝜌 is the fluid
density[kg/m3],Visthevolumeoftheimpingementzone[m3],𝑚isthemassflowrateofajet[kg/s],𝑢
isthevelocityofajet[m/s](withsubscripts1and2denotinginput1or2),disthediameter[m]andhis
theeffectivethicknessofthecylindricalmicromixingregion.Itisdifficulttoestimatewhatthecylindrical
micromixingregionwouldbegiventhenon-uniformshapeoftheliquidsheetafterimpingementandits
changingdimensionsdependingontheflowrate.IfdisassumedtobethejetdiameteroftheIJRwhile
the value of h to be 10% of the internozzle distance,147 mixing time can be estimated using the
correlationsuggestedbyCommengeandFalk:160
𝑡F = 0.15𝜖KL.CM (4.13)
where 𝑡F is the mixing time [s]. Using the correlation, the mixing time estimate was found to be
between0.9-2.6msforthe0.5mmIJRand0.25-0.9msforthe0.25mmIJRfortheconditionsusedin
thisstudy.ThemixingtimesfoundusingtheIEMmodelwerebetween1-6.5msforthe0.5mmIJRand
0.9-3.5msforthe0.25mmIJR.ConsideringthatboththecorrelationandtheIEMmodelonlygivean
estimateofthemixingtime,theorderofmagnitudeagreementissatisfactory.Increasingspecificpower
dissipationbyincreasingflowratewithintheflowreactorshoulddecreasemixingtime.Howeverthisis
notthecasefortheIJRwherethemixingtimelevelsoffandbeginstoincreasepastaWebernumberca.
90, most likely because of droplet formation and breakup at the liquid rim of the sheet formed on
impingementofthejets.ErniandElabaddiobservedthatthereisatransitionfromtheliquidsheettoa
fragmentedsheet (unstable rim)ataWe~30ata surfactant concentrationof8.2μMsodiumdodecyl
sulfate(SDS),We~80ataconcentrationof0.82mMandWe~200at82mM.149Theyobservedthatthe
mixing quality deteriorateswhen there is excessive droplet fragmentation; droplets ejected from the
sheethavelittletimeforefficientmasstransferwhichishighestintheliquidsheet(theyestimatethat
droplets took 0.6ms to be ejected from the liquid sheet).Using the fourthBourne reaction scheme,
Chapter4
77
theyestimatedmixingtimeintheirsystemtobe4.9ms,whichisofthesameorderofmagnitudeasour
work.The flowregime inoursystemwassubsequently investigatedtoascertainwhere this transition
occurs.
4.6 Flowregimecharacterizationofimpingingjetreactorusinghighspeedcamera
Theflowregimeofthe0.5mmand0.25mmIJRwasinvestigatedbycapturingimagesoftheIJRduring
operationusingahighspeedcamera(PhotronFastcamSA1.1highspeedcamera).Figure4-4showsthe
flowregime for the0.5mmIJRsystem, for flowratesbetween32-72ml/min (Webernumber13-65).
Fromtheimages,itcanbeseenthattheflowpatternchangesacrosstheflowraterangeinvestigated.
Atthelowestflowratethejetscollideandcoalesce,formingasmallsheet.Astheflowrateincreases,
thesheetincreasesinlengthanddiameter.Furtherincreaseintheflowrateresultsinawavyliquidfilm
surrounding the liquid sheet. This behavior has been observed in other studies where a wavy film
occurs,althoughitisusuallyaccompaniedbydropletsleavingthestablesheetboundedbyaliquidrim.
ThiseffectarisesfromRayleigh-Plateauinstability.164,165Itispossiblethatthesewaves,observedonthe
liquid rim and sheet, arise from small oscillations in the flowdelivered from the syringe pumps. This
wavy nature of the liquid sheet is also transferred into the chain of fluid following the initial sheet
formedbythecollisionofthejets.Theliquidfollowingthecollisionofthejetsremainscontinuousinall
cases for the0.5mm IJR,withnodropletbreakup in the regionvisualized.Figure4-5shows the flow
regimeforthe0.25mmIJRsystem,for flowratesbetween18-42ml/min(Webernumber32-176).At
lowerflowrates,astablesheetandliquidrimareformedatthepointofimpact.Atflowrateabove26
ml/min,anunstable rimwithdropletsbreakingoff themainstructurebegins to form.Fromthe front
viewitcanbeseenthattheflowisalsodevelopingasprayinthatplane.Atthispointtheinertialforces
overcome the surface tension forces of the fluid resulting in droplet breakaway from themain liquid
sheet/rim.Alsoforthe0.25mmIJRsystem,thechainoffluidfollowingtheliquidsheetsafterimpactis
non-continuousafterthefirstfewsecondaryliquidsheetsareformed.
Takingthemixingtimeestimationsandflowregimeobservationsintoaccount,themixingbehavior
ofthe IJRsystemcanberationalized.Therearetwomodesofatomization:reflectiveandtransmitive.
Between these two regions there exists a well-mixed atomization condition.166 In the reflective
atomizationregime,whichoccursatrelativelylowjetvelocities,thejetsessentiallycollideandcoalesce
forming a thin liquid sheet immediately after collision. Deflection of each jet will be equal as they
bounceoneachotherbecauseofequalmomentum. Inthis region, theturbulence in themixing layer
following impingement increases as flow rate increases leading to an increasing mixing efficiency.
However, because of the turbulent nature of the jets therewill be a fluctuation in the trajectory of
droplets/sheet as velocity increases; there is an eventual breakup of the liquid sheet where smaller
droplets break away. At this point the atomization transitions into a transmitive mode, whereby
dropletsareejectedfromtheliquidrimofthesheetbecauseofincreasedmomentum.Thisisobserved
inFigure4-5atWe>90.Thedropletsthatareejectedfromtheliquidrimhavelittlecontacttimeinthe
relativelyefficientlymixedsheetandthemixingefficiencysuffersbecausehigheramountsofdroplets
areejected.TheresultsshowthatbelowaWe~90,weareinthereflectiveatomizationregime.Above
thisWebernumbertheregimetransitionsintoatransmitiveregime.AWe~90seemstoyieldoptimal
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
78
mixingas confirmedby themixing timeestimation. Theobserved transition is somewhathigher than
thoseobservedbyErniandElabaddiwhoatminimalsurfactantconcentrationobservedtransitionatWe
~30.149Howevertheyuse jetsatanangleof110°witha jetspacingof10mmwhichcouldaffectthe
transitionpoint.
Figure4-4:Flowvisualizationof jetsemitted from0.5mmtubesat total flowrate:A:32ml/min (We:13),B:40
ml/min(We:20),C:48ml/min(We:29),D:56ml/min(We:39),E:64ml/min(We:51)andF:72ml/min(We:65).
Picturesshowsideviewandfrontview.
Chapter4
79
Figure4-5:Flowvisualizationof jetsemittedfrom0.25mmtubesattotalflowrate:A:18ml/min(We:32),B:22
ml/min(We:48),C:26ml/min(We:68),D:30ml/min(We:90),E:34ml/min(We:115),F:38ml/min(We:144)and
G:42ml/min(We:176).Picturesshowsideviewandfrontview.
4.7 InvestigatingtheeffectofwebernumberonNPmorphology
4.7.1 EffectofwebernumberonsilverNPsynthesis1(citratesystem)
TheeffectthattheWebernumberhasonthesilverNPswasinvestigatedinthe0.5mmand0.25mm
IJRsusingtrisodiumcitrateasastabilizingagent.Concentrationsofinput1were0.9mMsilvernitrate,6
mMtrisodiumcitrateandforinput2were1.8mMsodiumborohydride.Totalflowratesforthe0.5mm
IJRrangedfrom32ml/minto72ml/minandforthe0.25mmIJRrangedfrom18ml/minto34ml/min
(repeatabilityofthesynthesisisshowninAppendixC,FigureC1).
Figure4-6showsTEMimagesforNPssynthesizedinthe0.5mmIJRat32,48,64and72ml/min
wheretheaveragediameteranddispersityoftheNPsare7.9±5.8nm,7.7±5.0nm,5.0±2.8nmand
3.4±1.4nm.AveragediameterandpolydispersitydecreasewithincreasingWebernumber(flowrate).
T-testsshowedp-values<0.001whencomparingallthePSDsagainsteachother.Thissuggeststhatthe
PSDsareallstatisticallydifferent.
Figure4-7showsTEMimagesforNPssynthesizedinthe0.25mmIJRat20,24,28and32ml/min
wheretheaveragediameteranddispersityoftheNPsare6.4±3.4nm,5.9±2.1nm,5.0±2.5nmand
5.1±4.6nm.ItappearswithincreasingWebernumber(flowrate)theNPsreduceinaveragesizeupto
ca. 5 nm. T-tests showedp-values > 0.05when comparing thePSDsof 20 and24ml/min, 24 and32
ml/minand28and32ml/min(allotherp-valueswhere<0.001).ThesesuggestthatthePSDsarenot
statisticallydifferent;however,thisismostlikelydowntothelargepolydispersity.Visualanalysisofthe
TEMimagesrevealstwodifferentpopulationsofNPs,thosethataresmallerthan20nmandthosethat
arelarger.AlthoughthelargerNPswerepresentinsmallfrequencies(notlargerthan15countsforany
flow rate), there were enough present to affect the average size and dispersity calculations of NPs
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
80
synthesizedat20ml/minand32ml/min,butat24ml/minand28ml/minvirtuallyno largerNPsare
present. If NPs larger than 20 nm are removed from the t-test analysis, only NPs synthesized at 20
ml/minand24ml/minhaveap-value>0.05withallotherp-valuesbeing<0.001. FromthePSDs in
Figure4-7itcanbeseenthat,astheWebernumberisincreased,theaveragesizegetssmallerandthe
dispersionisreducedinthe<20nmrange.However,thepolydispersityishigheratthelargestWeber
numbereven though ithas the leastdispersecurve forNPs<20nm.This isbecausesynthesisat the
highestWebernumberproducedthehighestfrequencyoflargerNPs(>20nm),andthedisparityinsize
betweenthelargerandsmallerNPsishigherascomparedtoNPsproducedatlowerWebernumbers.
Figure4-8showsthepeakabsorbancevs.WebernumberobtainedfromUV-Visspectroscopyfor
the silverNPs obtainedusing the 0.5mm IJR (We= 13-65) and the 0.25mm IJR (We= 32-115). It is
worth noting that theUV signal did not change significantly 24 hours after synthesis (~1% change in
peak absorbance). For each jet diameter, the peak absorbance in general decreases with increasing
Weber number. Since the NP size identified from TEM images was primarily below 10 nm, the
decreasingpeakabsorbanceispossiblybecauseofadecreaseinsizeoftheNPs.ForsilverNPsthemolar
extinctioncoefficientincreaseswithsize,167andabsorbanceincreaseswithincreasingsizebelowthe10
nmrange,aspredictedbyMietheory.139Table4-1andTable4-2showtheFWHMandpeakwavelength
forsilverNPssynthesisedusing trisodiumcitrateasastabilisingagentatvarious flowrates in the0.5
mmand0.25mmIJRrespectively.TheFWHMforthe0.5mmIJRincreasessignificantlyafteraflowrate
48ml/minwhilethepeakwavelengthremainsfairlyconstant.ThereisagradualincreaseintheFWHM
with increasing flow rate for the 0.25mm IJR, while the peakwavelengths again stay approximately
constant.
TheTEMimagesandpeakabsorbanceoftheNPswouldsuggestthattheNPsdecreaseinsizewith
increasingWeber numberwhile the dispersity shows amore complex relationship. Firstly the results
obtainedfromthe0.25mmIJRwillbediscussed. IntermsoftheNPsynthesis, thesizeanddispersity
changes can be explained by taking into account mixing characteristics of the impinging jet at the
particularWebernumber.TheNPsbelow20nmarereferredtoassmallerNPsandthoseabove20nm
arereferredtoas largerNPs.AtlowWebernumbers,thesmallerNPstendtobeofalargerandmore
polydispersenature.AstheWebernumberincreases,theaveragesizeoftheseNPsdecreaseswhilethe
PSDsbecometighter.Thiscanbeexplainedthroughthemixingcharacteristicswithinthe liquidsheet,
whereincreasingturbulenceresultsinashortermixingtime.Borohydridehasastabilizingeffectonthe
NPs,136, 139, 145 and can also supply a high amount of electrons permole (up to amaximum of eight
electrons in theory).133-135Therefore, if there isefficientmixingwithinthe liquidsheet, therewillbea
sufficientsupplyofsodiumborohydrideinthelocalregionswherereactionistakingplace,resultingin
increasedstabilizationandsmallerNPs.Ifthereareinsufficientamountsofborohydridemolecules,less
stableNPs,whichcangrowrelativelyeasily,are formed.Poor localmixingefficiency,which results in
low local concentrations of borohydride,will lead to larger NPs, since borohydride can reducemany
times its own molar equivalent of silver ions. The formation of larger NPs can be compared to the
formation of iodine in the Villermaux-Dushmann reactions, which is caused by a local lack of
Chapter4
81
orthoborate ions. Hence, increasing themixing efficiency (Weber number) decreases the size of the
smallerNPs.
It issuggestedthatthe largerNPsareformedinsilvernitraterichregions.Thepresenceof large
NPswasmostfrequentataWebernumberof40and102.IntermediateWebernumbersof58and78
hadvirtuallynoNPsover20nm.Theappearanceof largerNPsandtheirchangeinsizeandfrequency
can be explained by mixing efficiency and the transition from a reflective to a transmitive type of
atomization. At the lowest Weber numbers, in the reflective atomization regime, there is lower
turbulencewithin themixing zone.This leads to thepossibilityof reactionsoccurring inpoorlymixed
regions,resulting intheformationof largerNPs.Atthe intermediateWebernumbers,veryfew larger
NPs are formed because of the increasing mixing efficiency. At large Weber numbers, there is a
transition into transmitiveatomization.At thispointanunstable rim is seenwithdroplets leaving the
mainbodyoffluid,resultinginanetdecreaseinmixingefficiency.Withinthesedroplets, it ispossible
thatlargerNPsareformedespeciallyiftheyaresilvernitraterich.
Thereductioninsizeanddispersityismoreobviousforthe0.5mmIJR.Thisreductionisobtained
as the mixing efficiency is increased within the liquid sheet, causing the average size of the NPs to
decrease andminimizing the formation of larger NPs. Again this is because the reactions are taking
placewithamoreefficientspreadofsodiumborohydride,whichincreasesstabilityofformedNPsand
decreasestheformationofNPsinsilvernitraterichregions.Thereisreductioninsizeandpolydispersity
whenthehydrodynamicsof the IJR transition fromauniform liquidsheet (We<29) toa liquidsheet
withwaves(We>39)withthesizeandpolydispersityultimatelyreducingfrom7.9±5.8nmto3.4±1.4
nm.
ForNPs less than10nm, thepeakwavelengthdoesn’tchangesignificantlywithsize.139This is in
linewiththeapproximatelyconstantpeakwavelengthsshowninTable4-1andTable4-2,andtheTEM
imagesshowNPswithanaveragediametersoflessthan10nm.TheincreasingFWHMcanpossiblybe
explained by the change in the larger NPs,where the size increases (particularly in the 0.25mm IJR
case).TheseNPswhicharelargerthan10nm,causeashiftinpeakwavelengthforthispopulationofthe
NPs,causingawideningoftheFWHM.Thisisonepossibleexplanationandtherecouldbesomeother
phenomenonwhichexplainsthewideningoftheFWHM.
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
82
Figure4-6:TEMimagesandparticlesizedistributionsofsilverNPssynthesisedatdifferenttotalflowratesusinga
0.5mmI.D.tubingIJR.(A)32ml/min(We:13),(B)48ml/min(We:29),(C)64ml/min(We:51)and(D)72ml/min
(We:65).Concentrationofsilvernitrate0.9mM,trisodiumcitrate6mMininput1andsodiumborohydride1.8mM
ininput2.
Chapter4
83
Figure4-7:TEMimagesandparticlesizedistributionsofsilverNPssynthesisedatdifferenttotalflowratesusinga
0.25mmI.D.tubingIJR.(A)20ml/min(We:40),(B)24ml/min(We:58),(C)28ml/min(We:78)and(D)32ml/min
(We:102).Concentrationofsilvernitrate0.9mM,trisodiumcitrate6mMin input1andsodiumborohydride1.8
mMininput2.
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
84
Figure4-8:Dependenceofpeakabsorbance(peakwavelength386-389nm)onWebernumberobtainedfromUV-
VisspectroscopyofsilverNPssynthesisedatdifferenttotalflowratesusing:i)0.5mm(blacksquares)andii)0.25
mm I.D. tubing IJR (red triangles).Concentrationof silvernitrate0.9mM, trisodiumcitrate6mM in input1and
sodiumborohydride1.8mMininput2.
Table 4-1: FWHMand peakwavelength of silverNPs synthesised in the 0.5mm IJR using trisodium citrate as a
stabilisingagentatflowratesbetween32and72ml/min.Concentrationofsilvernitrate0.9mM,trisodiumcitrate6
mMininput1andsodiumborohydride1.8mMininput2.
Flowrate(ml/min) WeberNumber FWHM(nm) Peakwavelength(nm)32 13 66 38740 20 66 38648 29 67 38656 39 73 38764 51 75 38772 65 77 387
Chapter4
85
Table4-2:FWHMandpeakwavelengthof silverNPs synthesised in the0.25mm IJRusing trisodiumcitrateasa
stabilisingagentatflowratesbetween18and34ml/min.Concentrationofsilvernitrate0.9mM,trisodiumcitrate6
mMininput1andsodiumborohydride1.8mMininput2.
Flowrate(ml/min) WeberNumber FWHM(nm) Peakwavelength(nm)18 129 72 38920 160 72 38922 193 74 38724 230 77 38726 270 78 38928 313 79 38930 360 87 38632 409 91 38834 462 89 388
4.7.2 EffectofwebernumberonsilverNPsynthesis2(PVAsystem)
TheWebernumbereffectonsilverNPswasalsoinvestigatedinthe0.5mmand0.25mmIJRsusingPVA
asastabilizingagent.Concentrationsofinput1were0.9mMsilvernitrate,0.02wt%PVAandforinput
2were1.8mMsodiumborohydride.Totalflowratesforthe0.5mmIJRrangedfrom32ml/minto72
ml/minandforthe0.25mmIJRrangedfrom18ml/minto42ml/min(repeatabilityofthesynthesisis
showninAppendixC,FigureC2).PVAasasurfactantaffectsthesurfacetensionoftheaqueoussolution.
Thesurfacetensionforwateris72mN/manditisestimatedthatPVAreducesittoavaluebetween60-
65mN/m.168Althoughthere isasmall increase inWebernumberbecauseof reducedsurface tension
usingPVA, thehydrodynamicsobservedshowedminimalchangewhencompared tousingwateronly
(seeAppendixC,FigureC3).
Figure4-9showsTEMimagesforNPssynthesizedinthe0.5mmIJRat32,48,64and72ml/min,
wheretheaveragediameteranddispersityoftheNPsare5.4±1.6nm,4.9±1.1nm,4.6±1.1nmand
4.2 ± 1.1 nm.Using this jet diameter, theNP size decreaseswith increasingWeber number (up to a
maximumofWe:65).T-testsyieldedp-values<0.001whencomparingthePSDs.Thedispersityofthe
NPs does not change significantly with increasing Weber number. One of the differences from the
trisodiumcitratestabilizedsystemisanotablelackoflargerNPs,resultinginlowerpolydispersity.This
difference can be attributed to the PVA molecules used as ligands which provide a bulkier steric
stabilization as opposed to the electrostatic stabilization of the smaller citratemolecule, leading to a
strongerbarriertopreventgrowththroughcoalescenceofNPs.
Figure4-10showsTEMimagesforNPssynthesizedinthe0.25mmIJRat18,26,34and42ml/min,
wheretheaveragediameteranddispersityoftheNPsare4.3±1.0nm,4.7±1.3nm,4.3±0.7nmand
4.6±1.1nm.ThesizeanddispersitydonotseemtobecorrelatedwiththeWebernumberinthiscase.
T-test yielded p-values of 0.68when comparing 18 and 34ml/min, 0.22when comparing 26 and 42
ml/min and < 0.001 in all other cases. It appears that in the 0.25mm IJR there is not a discernible
changeofNPsizewhencompared to the0.5mm jets,especiallywithp-valuessuggesting there isno
differencebetween18and34ml/minand26and42ml/min.
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
86
Figure4-11showsthepeakabsorbancevs.WebernumberobtainedfromUV-Visspectroscopyfor
thesilverNPsobtainedusingthe0.5mmIJR(We=13-65)andthe0.25mmIJR(We=32-176).Intermsof
stability,theUVsignalchangedinthefirst15minaftersynthesis,butremainedstableafter30minand
up to 2 hr. After 24 hr therewas a 16% increase in peak absorbance. For the 0.5mm IJR, the peak
absorbance decreases with increasing Weber number. For the 0.25 mm IJR, the peak absorbance
decreasesminimallyastheWebernumberincreases.Thisindicatesthatinthe0.5mmIJRcasetheNP
sizedecreaseswithincreasingWebernumber,whichisinagreementwiththeTEMobservations.Inthe
0.25 mm IJR case, the peak absorbance trend indicates that the NP size changes minimally with
increasingWebernumber,againinagreementwithTEMobservations.Table4-3andTable4-4showthe
FWHMandpeakwavelengthforsilverNPssynthesisedusingPVAasastabilisingagentatvariousflow
ratesinthe0.5mmand0.25mmIJRrespectively.Similartothetrisodiumcitratesystem,theFWHMfor
the 0.5mm IJR increases significantly after a flow rate 48ml/min and the peakwavelength remains
fairlyconstant.Again,similartothetrisodiumcitratesystem,thereisagradual increaseintheFWHM
withincreasingflowrateforthe0.25mmIJR,whilethepeakwavelengthstayapproximatelyconstant
exceptforthehighestflowratewhereitdecreasesby4nm.
PVA has a much higher molecular weight than trisodium citrate and is expected to be more
effective at stabilising the NPs through adsorption onto the silver NP surface through its hydroxyl
groups,whicharenumerousalongthechain.169, 170Citratemolecules,whicharemuchsmaller,adsorb
onthesurfaceoftheNPsthroughtheircarboxylgroups(citratehasthreecarboxylgroupsasopposedto
themanyhydroxylgroupsonaPVAchain).171Anotheraspectofthetwoligandsistheeffecttheyhave
onthekineticsofthereductionreactionandsubsequentnucleationofNPs.Itispossiblethattheligands
interact with the precursor in different ways; trisodium citrate behaves as a buffer and affects the
solutionpH,whereasPVAdoesnotbehave inthismanner.Literature indicatesthatthePVAmolecule
slows down growth of silverNPs compared to reductionwith no ligand present, by investigating the
evolutionofabsorbanceovertime.172Thisreductioningrowthrateismostlikelybecauseofthebulky
nature of the PVA, which would make it more difficult for clusters of silver to approach a growing
nucleus.Itissuggestedthat,sincetrisodiumcitratemoleculesaresmallerthanPVA,theywouldallowa
closerapproachforsilverclustersinsolutiontotheNPsurface(literaturesuggestsvariousclustersizes
suchasformationofAg13andAg2+).82,84Thepresenceofhighlyconcentratedsilvernitratezonesinthe
liquidusinganIJRinthecaseofcitratecappedNPswilllikelyresultingrowthatthesurfaceoftheNP
throughcoalescenceofsilverclusters.ThepossiblereasonthatnolargerNPsareseenusingthePVAare
that itprovidesamuch largerbarrierontheNPsurface, inhibitingtheapproachofsilverclustersand
rapid growth in areas that are rich in silver nitrate. At the same time, the stabilising effect of
borohydridecanbetakenintoaccount.Thereisabalancebetweenhowquicklyreducedsilverclusters
canapproachandincorporateontotheNPsurfacevs.theabundanceofborohydrideinthevicinityof
the NP. As mixing efficiency increases, there is a better distribution of borohydride, resulting in a
suppressedgrowthrateandsmallerNPs,whichistheobservedeffectusingPVAasa ligandinthe0.5
mmIJR.Usingthe0.25mmIJR,whichyieldsamoreconstantsizewithincreasingflowrate,thesizeof
theNPsdoesnotfallbelowca.4nmwithincreasingmixingefficiency.ThesizeoftheNPsusingthe0.25
Chapter4
87
mmIJRfluctuatesbetween4.3and4.7nm.Thereasonforthisfluctuationseemstobeduetotheflow
regimeinthe0.25mmIJR,withtheunstablerimregioncontributingtoafluctuatingsizewhichisnotas
repeatable as a stable rim. The repeatability data for the PVA system (Appendix C, Figure C2) also
indicates that silver NPs synthesized in the 0.25 mm IJR show more fluctuation in absorbance as
comparedtothe0.5mmIJR,furthersuggestingthattheunstablerimcontributestothefluctuatingNP
size in the 0.25 mm IJR system. It is expected that improved mixing efficiency should result in a
reductioninsizeasexplainedaboveandobservedinthe0.5mmIJR.Howeverthisreductioninsizeis
notseenwithimprovingmixingefficiencyinthe0.25mmIJRbecausethePVAmaybeimposingalimit
on the size of NPs. The synthesis was repeated in the 0.5 mm IJR at a We=65 using double the
concentrationofPVA(0.04wt%).Thesizewasreducedfrom4.2±1.1nmto3.7±1.0nm,suggesting
thatthePVAconcentrationisoneofthefactorsthatdeterminetheNPsize(TEMimageandPSDforthis
experimentareshowninAppendixC,FigureC4).ThissuggeststhattoachieveNPssmallerthan4nm,an
increaseinPVAwouldberequired,otherwiseonceacertainmixingefficiencyisachieved;thesizeofthe
NPdoesn’tdecreaseatafixedPVAconcentration.TheincreaseinFWHMinthecaseofboththe0.5mm
IJR and 0.25mm IJR is interesting, given the peakwavelength doesn’t change. The decrease in peak
absorbanceandconstantpeakwavelengthinconsistentwithareductioninsizeofNPsbelowthe10nm
size and the dispersity doesn’t seem to significantly increase according to the TEM analysis.
Interestingly,thesameeffectwasseenintermsoftheFWHMtrendwhenusingtrisodiumcitrateasa
stabilisingagentandthiswarrantsfurtherinvestigation.
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
88
Figure4-9:TEMimagesandparticlesizedistributionsofsilverNPssynthesizedatdifferenttotalflowratesusinga
0.5mmI.D.tubingIJR.(A)32ml/min(We:13),(B)48ml/min(We:29),(C)64ml/min(We:51)and(D)72ml/min
(We:65).Concentrationofsilvernitrate0.9mM,PVA0.02wt%ininput1andsodiumborohydride1.8mMininput
2.
Chapter4
89
Figure4-10:TEMimagesandparticlesizedistributionsofsilverNPssynthesizedatdifferenttotalflowratesusinga
0.25mmI.D.tubingIJR.(A)18ml/min(We:32),(B)26ml/min(We:68),(C)34ml/min(We:115)and(D)42ml/min
(We: 176). Concentrationof silver nitrate 0.9mM,PVA0.02wt% in input 1 and sodiumborohydride 1.8mM in
input2.
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
90
Figure4-11:Dependenceofpeakabsorbance(peakwavelength398-401nm)onWebernumberobtainedfromUV-
VisspectroscopyofsilverNPssynthesisedatdifferenttotalflowratesusing:i)0.5mm(blacksquares)andii)0.25
mm (red triangles) I.D. tubing IJR. Concentration of silver nitrate 0.9mM, PVA 0.02wt% in input 1 and sodium
borohydride1.8mMininput2.
Table4-3:FWHMandpeakwavelengthofsilverNPssynthesisedinthe0.5mmIJRusingPVAasastabilisingagent
at flow rates between 32 and 72ml/min. Concentration of silver nitrate 0.9mM, PVA 0.02wt% in input 1 and
sodiumborohydride1.8mMininput2.
Flowrate(ml/min) WeberNumber FWHM(nm) Peakwavelength(nm)32 13 80 40040 20 81 40148 29 82 39856 39 86 40064 51 91 40072 65 93 400
Chapter4
91
Table4-4:FWHMandpeakwavelengthofsilverNPssynthesisedinthe0.25mmIJRusingPVAasastabilisingagent
at flow rates between 18 and 34ml/min. Concentration of silver nitrate 0.9mM, PVA 0.02wt% in input 1 and
sodiumborohydride1.8mMininput2.
Flowrate(ml/min) WeberNumber FWHM(nm) Peakwavelength(nm)18 32 93 40122 48 95 40026 68 93 39930 90 97 40034 115 99 40038 144 101 39942 176 100 395
4.8 Conclusions
The impinging jet reactor (IJR) was characterized visually and using the Villermaux-Dushmann test
reactionsystemtoinvestigatetheflowregimeandmixingefficiencyunderarangeofflowratesandjet
diametersof0.25mmand0.5mm.AmaximummixingefficiencywasidentifiedataWebernumberof
ca.90,anditwasfoundtocorrespondtothepointatwhichtheflowregimetransitionsfromastable
rim sheet/chain like pattern to anunstable rimaccording to visual observationof the impinging jets.
Thispointisidentifiedasthetransitionfromareflectivetoatransmitivetypeatomization.
Two different ligands for silver NP synthesis were investigated using the IJR: trisodium citrate,
whichelectrostaticallystabilizesNPs,andPVA,whichisa largechainmoleculethatstericallystabilizes
NPs.TrisodiumcitrateproducedmorepolydispersesilverNPs.Averagesizewasreducedwithincreasing
flowratebecauseof increasedmixingefficiency,reducingthesilvernitraterichzonesoccurringinthe
mixingzoneafterimpingement.IncreasingmixingefficiencyinthePVAsystempastaWebernumberca.
60didnotyieldsmallersizeNPssuggestingalimitofsizeachievableusingPVAasaligand,possiblydue
tothesmallamountofPVApresent.MoremonodispersedNPsareobtainedusingPVA,indicatingthatit
ismoreefficientatstabilizingtheNPssinceithasmorepotentialsitesavailableforadsorptionontothe
surface of the NP than the citrate molecule, and it kinetically hinders growth during NP formation.
Figure 4-12 shows the diameter of the synthesised silver NPs against theWeber number for all the
synthesis conditions tested. The overall trend shows that theNP diameter decreaseswith increasing
WebernumberwhentheWebernumberisrelativelylow,andthenthedecreaseislesspronouncedand
levelsoffasWebernumber increasestorelativelyhighnumbers.Thetrendmatcheswellwiththatof
thedependenceofmixingtimeontheWebernumber,howeveritmustbenotedthatthetrendisalso
influenced by the surfactant that is used (PVA concentration appears to limit the size that can be
achievedinthatsynthesis).ThedecreasingsizeofNPwithincreasingnumberwasmoreapparentwhen
citratewas used as a surfactant. The lack of fouling and the efficientmixing in the IJR are themain
benefitsforthesynthesisofNPs.ThesizeoftheNPscanbereducedwith increasingmixingefficiency
while also reducing polydispersity. The IJR is scalable through a 'scale-out' procedure of essentially
increasingthenumberofIJRsoperatingintandem,makingitsuitableforlargescalesynthesisofNPs.
Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjetreactor
92
Figure4-12:DependenceofsilvernanoparticlediameterontheWebernumber.Datashownisfor0.5mmand0.25
mmIJRsandforbothcitrateandPVAsurfactants.
93
5 Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisof
goldandsilvernanoparticlesinacoaxialflowdevice
There aremany studies in the literature on the synthesis of gold and silverNPs. Twomore common
synthesismethodswereused inthisstudy:thewell-knownTurkevichsynthesisforgoldNPs88andthe
reductionofsilvernitrateviasodiumborohydrideinthepresenceoftrisodiumcitrateforsilverNPs.
TheTurkevichmethodforsynthesisinggoldNPsisperhapsthemostcommon,andtherearemany
studies in which batch reactors were used.89-106, 173 Flow syntheses of gold NPs using the Turkevich
methodare lesscommon.Ftounietal.synthesizedgoldNPswiththeTurkevichmethodusingafused
silicacapillaryandat-mixertomixchloroauricacidtosodiumcitratepriortointroducingthemintothe
heated capillary.107 Residence times within the capillary were varied between 35 to 94 s and they
obtained NPs in the size range between 1.5 and 3 nm, with larger NPs being obtained at longer
residence times. A minimum size of NPs was found at around 3.15 ratio of citrate to gold. The
temperatureeffectwas investigated,andNPsizedecreasedwith increasing temperature in the range
between60and100°C. Suganoetal. synthesizedgoldNPs in the size range10-45nmusinga y-type
micromixerwithapulsedflowofchloroauricacidandsodiumcitratetomixthereagentswithinasmall
channel at room temperature.108Mixing efficiencywas alteredby adjusting thepulsing rate between
50,100 and 200 Hz. The NP size increased with a higher pulsing frequency, suggesting faster mixing
resultedinlargerNPs.Wengetal.synthesizedgoldNPsusingtheTurkevichmethodrecipeusinganovel
microfluidicdevice.40HexagonalNPsofaround35nmwereobtainedata temperatureof115°Cwith
reaction time in the order of 2 to 5 min. There are also various studies in which gold NPs were
synthesizedwithmethodsotherthanthatofTurkevichinflowreactors.109-112,127
SilverNPsarecommonlysynthesizedinbatchwithsodiumborohydrideusedtoreducethesilver
precursor.5,64,66,67,69,70SARmixertypereactors71andPTFEchambers65havebeenutilizedtosynthesize
silverNPs inflowusingborohydride,obtainingNPs intherangeof10-20nmforborohydridetosilver
nitrateratiosrangingfrom3to40.WehavealsopreviouslyusedtheCFRandanimpingingjetreactor
forsynthesisofsilverNPsusingborohydrideasareducingagent.145,174
TheCFRhasbeenusedtosynthesizeavarietyofdifferentNPssuchastitania,128zirconia,129 iron
oxide,35, 36, 130 polymer,131, 175, 176 nickel,132 palladium177 and galliumnitride.178 The SARmixer uses the
principleofreducingthediffusiondistancebetweenmultiplestreamsthatarerequiredtomixandcan
becategorizedasaserial laminationdevice.SARmixershavebeenemployedforthesynthesisofgold
andsilverNPs.71, 109, 127Parallel laminationdevicesaresimilartoserial laminationbutthereduction in
diffusiondistanceisdoneinonestepratherthanmultipleones,andthesehavealsobeenusedforthe
synthesisofNPs.110,112TheCFIisprimarilyusedasaheattransferdeviceintheliterature,179,180andcan
beusedasaninlinemixer181buthasnotbeenusedspecificallyforthesynthesisofNPs.
Avarietyofdifferentconfigurationswereusedinthisstudy,withthefocalpointbeingtheCFR,for
the synthesis of gold and silver NPs were employed in this study. The CFR works on the basis of a
cylindricalinnerstreamofreagentwhichissheathedbyacylindricalouterstreamofreagent,creatinga
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
94
reactioninterfacebetweenthestreams.Theconfigurationsusedeither:addedanextraflowcomponent
following the CFR for a specific purpose or, by altering the hydrodynamics within the CFR. Flow
components that were used in conjunction with the CFR were: coiled flow inverter (CFI) for the
synthesis of gold NPs using a Turkevich recipe where tetrachlorauric acid and trisodium citrate are
mixed and heated, and a split and recombine (SAR) mixer for the synthesis of silver NPs through
reduction of silver nitrate by sodium borohydride in the presence of trisodium citrate. Schematic
representationsoftheseflowcomponentscanbeseeninFigure5-1.
Figure5-1:Schematicrepresentationsofcomponentsusedinthemicrofluidicflowsetup:a:coaxialflowreactor,b:
coiledflowinverterandc:splitandrecombinemixer.
TheCFIwasusedasaresidenceloopfollowingtheCFRbecauseofitsmixingcharacteristics.Ithasideal
propertiesinsynthesizingNPsbecauseofanimprovedresidencetimedistribution.182Thisimprovement
arises from enhancedmass transfer because of a secondary flow developingwithin a helically coiled
channel,knownas‘Deanflow’.ThisischaracterizedbytheDeannumber,whichistheratiobetweenthe
squarerootoftheproductofinertialandcentripetalforcesoverviscousforces:
De= d2r 5udN= d
2rRe
wheredisthediameterofthechannel,ristheradiusofcurvatureofthecoil,𝜌isthefluiddensity,uis
themeanvelocityofthefluid,𝜇istheviscosityofthefluidandReistheReynoldsnumber,theratioof
inertialtoviscousforces.Themixingisfurtherimprovedwhenthissecondaryflowisinverted,asinthe
CFI,byintroducing90°bendsinthehelicallycoiledchannel.TheSARisadevicewhichimprovesmixing
byreducingthediffusiondistancebetweenstreamsbycontinuouslysplittingandrecombiningtheflow
Chapter5
95
intothinnerandthinnerinterdigitatedlamellae.HydrodynamicchangesintheCFRwhenoneoperates
atahigherRenumberasopposedtolaminarflowbyincreasingthevelocityoftheinnerstreamorwhen
onechangesthethicknessoftheinnerstreamtoreducethediffusiondistance.Thevariouscomponents
andhydrodynamicconditionswerechoseninanefforttodemonstratewhattheeffectofmasstransfer
isonthesynthesisofgoldandsilverNPsintheCFR.
NPs were synthesized under a variety of different hydrodynamics and mixing conditions to
understandhowthesecouldaffectsizeanddispersity.Asmentionedpreviously,understandinghowto
controlsizeanddispersityisimportantbecauseitallowsNPstobetailoredforspecificapplications.Two
systemswerestudiedinthiswork:thesynthesisofgoldNPsintheCFRinconjunctionwiththeCFIusing
the Turkevichmethod recipe and the synthesis of silver NPs in the CFR in conjunctionwith the SAR
mixer,aswellastheCFRinisolation,usingsodiumborohydridetoreducesilvernitrateinthepresence
of trisodium citrate. The effect of flow rate (or residence time) and temperature on NP size and
dispersitywere investigated for thegoldNP system.Theeffectof silver concentrationon theNP size
and dispersitywas investigated using the CFRwith SARmixer system.Using the CFR in isolation, the
silverNPsynthesiswasperformedathighRe,toinvestigatetheeffectofhydrodynamicssilverNPsize
and dispersity, and the effect of changing the internal diameter of the inner tube in a laminar flow
regime(toreducethediffusiondistanceoftheinnerstream)onNPsizeanddispersitywasinvestigated.
5.1 Experimental
5.1.1 Chemicals
Silvernitrate(AgNO3,0.01Mstocksolution),trisodiumcitrate(HOC(COONa)(CH2COONa)2·2H2O,powder
form,sodiumborohydridesolution(NaBH4,~12wt%in14MNaOHstocksolution)andgold(III)chloride
hydrate (HAuCl4.xH2O, powder form) were obtained from Sigma. All chemicals were used without
furtherpurificationandsolutionswerepreparedwithultrapurewater(resistivity15.0MΩ·cm).
5.1.2 ExperimentalSetup
Syringepumps(Pump11EliteOEMmodule,Harvard)wereusedtodeliverthereagents.Thesetupfor
goldNP synthesis using theCFR followedby aCFI acting as a residence time loop, as canbe seen in
Figure5-2.TheCFRconsistedofa0.798mminternaldiameter(I.D.)innerglasstube(1.09mmexternal
diameter) and a 2mm I.D. outer glass tubewith the length from theoutlet of the inner tube to the
outletoftheCFRof21mm.Upstream,thecapillarieswereconnectedusinganETFET-piececonnector
(0.5mmthru-hole,UpchurchScientific)whichwasdrilledtoadiameterof2mmtoallowtheinnertube
tobeinsertedintotheoutertube.Theinnertubewasinsertedintoa1mmI.D.PTFEtubesleeve(1.59
mmexternaldiameter)tokeepthetubestableswithinthelargeroutertube.Downstream,theCFRwas
connectedtotheCFIbyusinganETFEunion(0.75mmthru-hole,UpchurchScientific).TheCFIwasa3.5
m 1mm I.D. PTFE tube that had nineteen 90° bendswith 20 arms, each arm consisting of 5.5 coils
(lengthalsoincludesentranceandexitlengthoftubing).Thecoilshadanaxialpitchof3mm,whilethe
coil diameterwas 8mm. The coilswere fabricatedbydrilling 1.6mmholes (using aRoland EGX-400
engravingmachine)intoaPTFEplate(2mmthickness,DirectPlastics)andthePTFEtubingwasthreaded
throughtheseholeswhichwerearrangedtoresultincoilingwiththedimensionsstatedabove.TheCFR
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
96
andCFIwere immersed ina glycerolbath (GR150,Grant Instruments)with temperaturebeing varied
between 60-100°C. The tetrachloroauric acidwas fed through a preheating CFI consisting of four 90°
bendsandsimilaraxialpitchandcoildiameterastheCFIdescribedabovetoensureitcameintocontact
with the trisodium citrate at the desired temperature. The outlet of the CFI was split to a waste
container and a sample container. These containers were pressurized using nitrogen gas and a
backpressureregulator,whichwassetto2barstopreventvaporizationofwaterwithinthereactor.
ThesetupsusedforthesynthesisofsilverNPscanbeseeninFigure5-3.TheCFRusedwassimilar
tothatinthegoldNPsynthesis,butthelengthfromtheinnertubeoutlettotheoutletoftheCFRwas
130mm (a longer lengthwas used for additional residence time)while the inner tube I.D. usedwas
between0.142mmand0.798mm (between0.559mmand1.09mmexternal diameter). A glass SAR
mixer(Micromixerchip,Dolomitemicrofluidics)wasplaceddownstreamoftheCFRinordertoenhance
the mixing by laminating the stream into thinner striations. The SAR mixer had an internal channel
diameterof0.125mmx0.35mm(depthxwidth)ofthemainchanneland0.05mmx0.125mmofthe
secondary channels responsible for splitting the flow. The internal volume was 8 μl. The CFR was
connectedtotheSARusinganETFEunion(0.75mmthru-hole,UpchurchScientific).FollowingtheSAR,
a10cmlong1mmI.D.PTFEtubedeliveredthesilverNPstothesamplecontainer.
Figure5-2:SchematicrepresentationofexperimentalsetupforgoldNPsynthesis.
Chapter5
97
Figure5-3:Schematic representationexperimental setup for silverNP synthesis: a)CFRused in conjunctionwith
SARandb)CFRusedinisolation.
5.1.3 Nanoparticlesynthesis
For the gold NP synthesis, initially the reactor setupwas filled with tetrachloroauric acid before the
systemwaspressurizedusingnitrogengasandabackpressureregulatorsetto2bar.Thereactorsetup
was then immersed in the temperature bath (60-100°C). A trisodium citrate solution and
tetrachloroauricacidweredilutedtotheappropriateconcentrationsandpumpedthroughtheinnerand
outer streams of the CFR respectively. Itwas found that filling the reactorwith tetrachloroauric acid
prior to synthesis reduced the amount of fouling drastically. The reactor was then operated at the
appropriate flowrate,andone reactorvolumewasallowed topass throughathigh temperatureand
collectedinthewastecontainerbeforetheoutletoftheCFIwasswitchedtosamplecollectiontocollect
10mlofsample.Smallamountsoffoulingwereobservedwhenthereactorwasoperatedataflowrate
of 0.25ml/min towards the end of the CFI. Itwas suspected this is because the long residence time
enabled citrate to diffuse towards the channelwallswhichwould increasepH. The surface chargeof
PTFEisclosetozeroatpH3.5,whichiscoincidentallyaroundthepHofthetetrachloroauricacid.183This
wasthereasonforthereductioninfouling,sinceareactorfilledwithamixtureofcitrateandprecursor
prior to immersion in the temperature bath resulted in increased fouling at the channel walls. The
surfacechargewouldbenegativeinthepHrangenearneutral(whichisthepHwhenmixingcitrateand
tetrachlorauricacidattheseconcentrations),whichprobablyattractsNPsinthegrowthphasebecause
ofthepositiveAuionadsorbedontothesurfaceoftheNP.Thusitisbeneficialtoallowasmalllayerof
tetrachloroauricacidtoremainatthewalltokeepthesurfacechargeneutral.Inlaminarflowthelayer
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
98
willremainforsometimebeforecitratediffusesnearthewallbecauseoftheflowprofileandnon-slip
conditionatthewallsofthereactor.Thisisevidencedbysmallamountsoffoulingonlyoccurringat0.25
ml/minwheretherelativeresidencetimeislonger.
For the silver NP synthesis, silver nitrate and trisodium citrate solutions were premixed to the
appropriate concentration and pumped through the inner stream. Sodium borohydride/sodium
hydroxidesolutionwasdilutedtotheappropriateconcentrationandpumpedthroughtheouterstream
of the CFR. Sodium borohydride was stored in 14M sodium hydroxide; hence the concentration of
sodiumhydroxidewas3.21timeshigherthanthestatedsodiumborohydrideconcentrationinallcases.
SilverNPsyntheseswerecarriedoutatroomtemperature(22-24°C).Allconcentrationsarethoseatthe
inletsbeforeanymixingofreagentsoccurs,unlessstatedotherwise.
5.1.4 Characterizationofnanoparticles
NPs were analysed using a UV-Vis spectrometer (USB 2000+ Spectrometer and DT-Mini-2-GS light
source,OceanOptics).SilverNPswereanalysedwithinanhourofsynthesis(thesignalofthesamples
were stable in thiswindowof time).GoldNPswere analysedover time andwere analysedonce the
signal was stable (the signal became stable usually after 1-2 days). NP samples were diluted with
additionalultrapurewatertobringtheabsorbanceintoasuitablerange(i.e.obeyingtheBeer-Lambert
lawandavoidingsaturationofthelightdetector)ifnecessaryandthedatawerenormalizedsothatthe
maximum absorbance in the particular set of experiments tested represented an absorbance of 1.
TransmissionelectronmicroscopeimageswerecapturedusingaJEOL1200EXiimicroscopewitha120
kVaccelerationvoltage.CarboncoatedcopperTEMgridswerepreparedwithinanhourofsynthesisfor
silverNPsandwhenthesamplewasstable forgoldNPsbypipettinga5µl sampleonto thegridand
allowing it todryatroomtemperature.Whenpreparingthesample, thedropletplacedonthegrid is
smalltoallowforquickerdryingandfortheringofaggregatedNPstobeidentifiedeasily(asdiscussed
in Chapter 3). Particle size distributions (insets for each TEM image presented) have the following
nomenclature:d is averagediameter,δd is the standarddeviationof theNPdistributionandn is the
number of particles counted to obtain the particle size distribution. Differential centrifugal
sedimentation analysis (CPS disc centrifuge UHR, Analytik) was carried out on gold NPs when the
samples were stable. PEBBLES software was used for the counting and sizing of TEM images of the
synthesised silver NPs. ImageJ software was used for the counting and sizing of TEM images of the
synthesisedgoldNPs.PEBBLESsoftwaretendstobebetterforlowercontrastimageswheretheoutlines
oftheNPsarehardertodistinguish,inthiscaseitwasmoresuitabletousethisprogramforsilverNPs.
TheImageJsoftwarecancountandsizeNPsfasterthanPEBBLES,butworksbetterwithhighercontrast
images.GoldNPimagesweresuitablyhighcontrasttoobtainaccuratesizedistributionsusingImageJ.
5.2 Resultsanddiscussion
5.2.1 Effectofincreasingflowrateongoldnanoparticlesizeanddispersity
TheeffectthattheflowratehasonthegoldNPswasinvestigatedintheCFRwithaninnertubeinternal
I.D.of0.798mmandanouter tube I.D.of2mm.The concentrationof trisodiumcitratewas0.09M
Chapter5
99
throughtheinnertube;theconcentrationoftetrachlorauricacidwas0.557mMthroughtheoutertube.
Thevolumetricflowrateratiowasfixedat32.3:1(Qout:Qin)andthetotalflowratewasvariedbetween
0.25and3ml/min.Themolarflowrateratiowas1:5(HAuCl4:Na3citrate).TheReynoldsnumbersvaried
between3and32intheCFRandbetween5and63intheCFI.TheDeannumberswererangedfrom1.9
to22.4intheCFIwhichisabovethethresholdof1.5requiredforfullydevelopedDeanflow.182
Figure5-4showsTEMimagesofgoldNPssynthesizedataflowrateof0.25,0.5,0.75,1,1.5,2and
3ml/minwheretheaveragediameteranddispersityoftheNPsare18.9±2.3nm,18.4±2.8nm,19.7±
2.4nm,17.9±2.1nm,20.6±7.2nm,21.7±3.3nmand23.9±4.7nmrespectively.Figure5-5shows
correspondingDCSmeasurements,wheretheaveragediameteris16.7±2.8nm,15.1±3.6nm,16.4±
3.0nm,14.9±2.3nm,18.5±3.1nm,22.5±3.8nmand26.4±4.0nmnmrespectively.Thetrendofthe
averagediameterof theNPsmatcheswith thatobserved fromTEM images. Figure5-6 showsUV-Vis
spectraofgoldNPssynthesizedataflowrateof0.25,0.5,0.75,1,1.5,2and3ml/min.Experimentsat
selected flow rateswere repeated three times and analyzedusingDCS (shown inAppendixD, Figure
D1).TheaveragesizeoftheNPsfrom0.25ml/minto1ml/minchangesminimallyaccordingtotheTEM
andDCSanalysis(althought-testsshowthatallPSDsarestatisticallydifferenttoasignificancevalueof
at least 0.05). Themolar extinction coefficient increaseswith size for goldNPs,184, 185 therefore for a
constantnumberconcentrationofNPs,anincreaseinpeakabsorbancewouldindicateincreasingsize.
However, the number concentration of NPs of changing size would change for a constant gold
concentration i.e. largerNPssynthesizedat thesamegoldconcentrationwouldhavea lowernumber
concentration.Hendeletal.showedthattheabsorbanceat400nmofcitratecappedgoldNPsinthe15-
30nmrangeincreaseslinearlywithincreasingsizeforaconstantgoldconcentration.186Thissuggestsa
largersizehasamoresignificanteffectontheabsorbancethennumberconcentrationofNPs.Figure5-6
showsthatthereisanincreaseintheabsorbanceat400nmwithincreasingflowratebutitisunlikely
thatthisisbecauseofincreasingsizesinceTEMandDCSshowamoreconstantsizeinthe0.25-1ml/min
range. The peak wavelength should redshift with increasing size,184 but the wavelength changes
minimally between 523 - 525 nm further suggesting there is aminimal change in size. One possible
explanation for the increase in absorbance up to 1 ml/min may be related to fouling. Fouling was
observed at the lowest flow rate suggesting an increase of fouling with decreasing flow rate. This
explainsthedecreaseinabsorbanceatlowerflowrates(despiteaconstantNPsize)sincesomeofthe
goldremainswithinthereactor.Increasingtheflowratefrom1.5ml/minto3ml/minshowsanincrease
in the size of the NPs from 20.6 to 23.9 nm according to TEM analysis, and from 18.5 to 26.4 nm
accordingtoDCSanalysis.Thepeakabsorbanceincreaseswithincreasingflowrateinthisrangeandthe
peak wavelength increases from 525 to 529 nm, which indicates an increase in the size of the NPs
consistentwithDCSandTEMdata.
The data shows that the size stays fairly constant between 0.25 and 1 ml/min and thereafter
increasesuptoaflowrateof3ml/min.Thiscanberationalizedasfollows.IntheCFRsection,thereisa
laminarflowwheretheouterflowoftetrachloroauricacidfocusesthe innerflowoftrisodiumcitrate.
Thereislittlemixinginthisregion.Masstransferoccursthroughdiffusionattheinterfacebetweenthe
twostreams.Inthemoreacidicconditionsofthetetrachloroauricacidstream,thereactionrateisfaster
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
100
andnucleationisfavored.Inthemorebasicconditionsinthetrisodiumcitratestream,thereactionrate
is slower. This is because of the speciation of the precursor, in acidic conditions the speciation is
favoured towards AuCl4-,95, 102, 104 whereas in more basic conditions the speciation passes through
various forms fromAuCl3OH- toAuOH4.As thegoldspeciesacquiremorehydroxyl ions, the reactivity
decreasesandthere isatransitionfromnucleationofgoldatomsintoclusterstogrowthofgoldonto
existing clusters. In the CFR region of the reactor, the tetrachloroauric acid and the trisodium citrate
remainwell separatedandonlyexchangematerial throughdiffusionat the interface. In this region,a
higherreactionrateandhencenucleationwilloccurinthetetrachloroauricacidstream.Therefore,the
CFRbehavesasanucleationsection.FurtheralongintheCFIresidenceloop,thestreamsaresubjected
toDeanflowandflowinversion.Theseenhancethemixingandlowerthereductionrateofgold ions,
hencetransitioningfromthenucleationsectionintheCFRtoagrowthsectionintheCFI.Giventhis,the
longernucleationcanbemaintained, the smaller the resultantNPs shouldbeobtained. In this study,
thereisamarkedincreaseinthesizeoftheNPsastheflowrateincreasespast1ml/min,supportingthe
hypothesisofnucleationdominatingintheCFR,sinceathigherflowrates,reactionsattheinterfaceof
streamsdonotpersistaslong.Thereisalimitationonsizereductionwhenoperatingbelow1ml/min,
which may be related to a decrease in citrate concentration. The decrease in citrate concentration
wouldresultinadecreaseinreductionofgoldprecursor,leadingtoasuppressionofnucleationwithin
theCFRsection.Finally,thePecletnumberforthesyntheseswas>>1atallflowrates, indicatingthat
masstransportwasadvectiondominated(Table5-1showsPecletnumberfortheCFRsectionwherethe
I.D.was2mmandthediffusioncoefficientusedwas1.4x10-9m2/s,187whichisfortetrachlorauricacid,
the Peclet number is doubled for those shown in the table for the CFI where the I.D. was 1 mm).
IncreasingPecletnumber indicateddecreasingsignificanceofdiffusionontheprocess, soagreeswith
theideasdiscussedabovetoexplainwhythegoldNPsizeincreaseswithincreasingflowrate.
Chapter5
101
Figure5-4:TEMimagesofgoldNPssynthesizedusingtheCFRwithaCFI residence loopata flowrateofA:0.25
ml/min ,B: 0.5 ml/min, C: 0.75 ml/min, D: 1 ml/min, E: 1.5 ml/min, F:2 ml/min and G: 3 ml/min. H: Average
diameter of goldNPs vs flow rate of synthesis (bars represent standard deviation of the size). Concentration of
tetrachloroauric acid, 0.557mM; concentration of trisodium citrate, 0.09M; volumetric flow rate ratio, 32.3: 1
(Qout:Qin,HAuCl4:Na3citrate);molarflowrateratio,1:5(HAuCl4:Na3citrate);temperature,80°C.
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
102
Figure5-5:NormalizedDCScurvesofgoldNPssynthesizedusingtheCFRwithaCFIresidence loopat flow rates
between 0.25 and 3 ml/min. Inset: Diameter of gold NPs vs. flow rate of synthesis (bars represent standard
deviationofsize).Concentrationoftetrachloroauricacidwas0.557mMandconcentrationoftrisodiumcitratewas
0.09M.Concentrationoftetrachloroauricacid,0.557mM;concentrationoftrisodiumcitrate,0.09M;volumetric
flowrate ratio,32.3:1 (Qout:Qin,HAuCl4:Na3citrate);molar flowrate ratio,1:5 (HAuCl4:Na3citrate); temperature,
80°C.
Figure5-6:UV-VisspectraofgoldNPssynthesizedusingtheCFRwithaCFIresidence loopatflowratesbetween
0.25and3ml/min.Inset:Peakabsorbance(blacksquares)andpeakwavelength(reddiamonds)ofgoldNPsvs.flow
rate of synthesis analyzed using UV-Vis spectroscopy. Concentration of tetrachloroauric acid, 0.557 mM;
concentrationoftrisodiumcitrate,0.09M;volumetric flowrateratio,32.3:1 (Qout:Qin,HAuCl4:Na3citrate);molar
flowrateratio,1:5(HAuCl4:Na3citrate);temperature,80°C.
Chapter5
103
Table5-1:DiameteranddispersityofgoldNPssynthesisedatvariousflowrateswithcorrespondingPecletnumber
Flowrate(ml/min) PecletNumber Diameter(nm) Dispersity(nm)
0.25 1.9x103 18.9 ±2.3
0.5 7.6x103 18.4 ±2.8
0.75 11.4x103 19.7 ±2.4
1 15.2x103 17.9 ±2.1
1.5 22.7x103 20.6 ±7.2
2 30.3x103 21.7 ±3.3
3 45.5x103 23.9 ±4.7
5.2.2 Effectoftemperatureongoldnanoparticlesizeanddispersity
Theeffect that the temperaturehasongoldNPs synthesiswas investigated in theCFRwithan inner
tubeinternalI.D.of0.798mmandanoutertubeinternalI.D.of2mm.Theconcentrationoftrisodium
citratewas0.09M through the inner tube,while the concentrationof tetrachlorauric acidwas0.557
mMthroughtheoutertube.Thevolumetricflowrateratiowasfixed32.3:1(Qout:Qin)andthetotalflow
ratewasfixedat1ml/min.Themolarflowrateratiowas1:5(HAuCl4:Na3citrate).TheReynoldsnumber
intheCFRwas11andintheCFIwas21.IntheCFI,theDeannumberwas7.5.
Figure5-7showsTEMimagesofgoldNPssynthesizedatatemperatureof60,70,80,90and100°C
wheretheaveragediameteranddispersityoftheNPsare25.5±5.7nm,22.7±3.9nm,17.9±2.1nm,
18.6±2.1nmand19.4±2.4nm respectively. T-tests for thePSDs showedp-valuesof<0.001when
comparing all PSDs, indicating statistical difference to a significance level of 0.001. Figure 5-8 shows
correspondingDCSmeasurements,wheretheaveragediameteris31.1±5.5nm,18.4±3.8nm,14.9±
2.3nm,16.2±3.4nmand17.6±2.7nmrespectively.The trendof theaveragediameterof theNPs
matchesthatobservedintheTEMimages.Figure5-9showscorrespondingUV-Visspectra.
TheaveragesizeoftheNPsdecreasedfrom25.5nmto17.9nmwhenincreasingthetemperature
from60 to80°Cand then increased slightly to19.4nmwhen increasing the temperature from80 to
100°C. DCS characterization also shows similar trends. The peak absorbance is highest at 70°C after
whichitdecreaseswithincreasingtemperaturewhilethepeakwavelengthchangesfrom528to523nm
from60to80°Candfrom523to525nmfrom80to100°C.Thecharacterizationconfirmsthatthereisa
minimumat80°C.Turkevichetal.alsoobservedaminimumsizeof16.5nmat80°C.88ChowandZukoski
found thatNPshadapproximately the same size at synthesis temperaturesof 60, 70 and80°Cbut it
took longer for NPs at lower temperatures to arrive at the final size.91 Wuithschick et al. found a
minimum of NP size at 60°C using the Turkevich method,104 while Piella et al. found a minimum at
70°C.105 Increasing temperatureaffectsanumberof factors in thesynthesis.TheconversionofAuCl4-
speciestohydroxylatedspeciesspeedsupathighertemperaturesmeaningthatnucleationperiodalso
terminatesfasterasthemorereactivespecies isconvertedtolessreactivespecies.Notealsothatthe
equilibriumshiftstomorehydroxylatedformsathighertemperatures.104However,therateofreduction
ofAuCl4- athighertemperatureswouldalsospeedup,resultinginahighernucleationrateinashorter
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
104
timeperiod. Finally, the sizeof the stablenucleiwould alsobe slightly larger at higher temperatures
becauseofthelargerthermalenergyavailable.104Thus,thereisabalanceofcompetingeffects,resulting
inaminimumsizeatanintermediatetemperature.Afurtherpointtonoteisthattheheattransferand
mixingwouldalsoplaya role inhowthe temperatureaffects theresultantNPs.For instance inbatch
synthesis, the citrate is injected at room temperature to boiling citrate, and so, depending on the
amountandnatureoftheinjection,thereactioncouldtakeplaceatdifferenttemperaturesinlocalized
regionsaroundtheinjectionpoint,whichmayexplainthevarietyofresultsreportedintheliterature.
The repeatability of the synthesis was investigated by repeating the experiment at each
temperature three times and characterizing the size using DCS (Appendix D, Figure D2). The
repeatability in sizewas found to beworst at 60°C and improvedwith increasing temperature up to
80°C,afterwhichitremainedfairlyconstant.
Chapter5
105
Figure5-7:TEMimagesofgoldNPssynthesizedusingtheCFRwithaCFIresidenceloopatatemperatureofA:60°C,
B:70°C,C:80°C,D:90°C,E:100°C. F:Averagediameterof goldNPsvs temperatureof synthesis (bars represent
standard deviation of the size). Concentration of tetrachloroauric acid, 0.557 mM; concentration of trisodium
citrate,0.09M;volumetricflowrateratio,32.3:1(Qout:Qin,HAuCl4:Na3citrate);molarflowrateratio,1:5(HAuCl4:
Na3citrate);flowrate,1ml/min.
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
106
Figure5-8:NormalizedDCScurvesofgoldNPssynthesizedusingtheCFRwithaCFIresidenceloopattemperatures
between60and100°C.Inset:DiameterofgoldNPsvs.temperatureofsynthesis(barsrepresentstandarddeviation
ofsize).Concentrationoftetrachloroauricacid,0.557mM;concentrationoftrisodiumcitrate,0.09M;volumetric
flow rate ratio, 32.3: 1 (Qout:Qin, HAuCl4: Na3citrate);molar flow rate ratio, 1:5 (HAuCl4: Na3citrate); flow rate, 1
ml/min.
Figure5-9:UV-VisspectraofgoldNPssynthesizedusingtheCFRwithaCFIresidenceloopattemperaturesbetween
60 and 100°C. Inset: Peak absorbance (black squares) and peak wavelength (red diamonds) of gold NPs vs.
temperature of synthesis. Concentration of tetrachloroauric acid, 0.557mM; concentration of trisodium citrate,
0.09 M; volumetric flow rate ratio, 32.3: 1 (Qout:Qin, HAuCl4: Na3citrate); molar flow rate ratio, 1:5 (HAuCl4:
Na3citrate);flowrate,1ml/min.
Chapter5
107
5.2.3 Silver nanoparticle synthesis usinga coaxial flowdevice followedby a split and recombine
mixer
ASARmixerwasattachedontotheCFRtoseetheeffectthatmulti-laminationwouldhaveonthesilver
NPsynthesis.TheCFRhadaninnertubeI.D.of0.556mmandanoutertubeI.D.of2mm.Thedistance
from the inner tube outlet to the outlet of the CFR was 130 mm. The effect that the silver nitrate
concentration had on the silver NPs was investigated using this configuration, previously we have
studiedtheeffectofsilvernitrateconcentrationusingtheCFRinisolation.145Silvernitrateandtrisodium
citrate solutionswere premixed before pumping through the inner tubewith concentration of silver
nitrate varied between 0.05 and 0.4 mM, the trisodium citrate concentration was fixed at 0.5 mM.
Concentrationof sodiumborohydridewas fixedat0.3mMandpumped through theouter tube. The
volumetric flow rate ratiowas fixed at 1:1 (Qout:Qin) and the total flow ratewas fixed at 2.5ml/min.
Experimentswerecarriedoutatroomtemperaturebetween22-24°C.
Figure5-10showsTEMimagesofsilverNPssynthesizedatasilvernitrateconcentrationof0.05,
0.15,0.25and0.4mM.TheaveragediameteranddispersityoftheNPswere5.5±2.4nm,4.7±2.8nm,
4.4±1.8nmand3.4±1.4nm.T-testsforthePSDsshowedp-valuesof<0.001whencomparingallPSDs
exceptwhencomparing0.15and0.25mMwherethep-valuewas0.15.Inthiscasethereisnostatistical
differencebetween0.15mMand0.25mMaccordingtothep-value.Figure5-11showscorresponding
UV-Vis spectra. Peak absorbance increases with increasing concentration, the peakwavelength stays
relativelyconstantbetween398-400nmbetween0.05-0.25mM,howeverat0.3mMthewavelength
decreases to395nmandat0.4mM thewavelength is 392nm.Althoughpeakwavelength shouldn’t
change according to Mie theory for silver NPs less than 10 nm,139 a blue shift in peak wavelength
generallyindicatessmallerNPs.
ThelaminarflowprofileinsidetheCFRcreatesaninterfacebetweenthesilvernitrateandsodium
borohydride,resultinginmixingthroughdiffusionratherthanenhancedbyastretchingandthinningof
the lamellae though turbulent flow for example. The laminar flow profile prevents reaction in silver
nitrate rich zones resulting in smaller NPs. Larger NPs occur in silver nitrate rich zones because the
sodium borohydride can reduce multiples of its own stoichiometric equivalent in silver ions,133-135
resultinginlargerandlessstableNPswhichareabletogrow.ThroughtheSARmixerthelaminarflow
conditionismaintainedbutthemixingefficiencyisincreasedsincethediffusiondistanceisreducedas
thestreamiscontinuouslysplitandrecombinedinawaytoseriallylaminatetheflowintothinnerand
thinnermultiplesof theoriginal stream,aswell as the creationof an increasednumberof interfaces
between reagents. In the experiments above, increasing silver nitrate concentration, led to higher
supersaturation levels and thus a higher nucleation rate resulting in smaller NPs. In previous
experimentswehaveconductedusingonlythecoaxialflowreactor,theoppositetrendwasobserved;
NPsizeincreasedfrom3.7±0.8nmto9.3±3nmwithincreasingsilverconcentrationfrom0.05-0.4mM
with the same sodium borohydride and trisodium citrate concentrations (a comparison is shown in
Figure5-12).145ThisoccurredbecausereactionwasnotcompleteintheCFRandtheadvectivemixingin
thedroplets at theendof theCFRallowed thepossibility of sodiumborohydride to travel into silver
nitraterichzones,whichasdiscussedpreviously,wouldleadtolargerNPs.Sincedownstreammixingis
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
108
morecontrolledwiththeSARmixer,silvernitraterichzoneswhichgiverisetolargerNPsarelesslikely
to occur. The above indicates that the mixing condition of the fluid at the exit of the CFR has a
pronouncedeffectonNPsynthesis.
Figure 5-10: TEM images of silver NPs synthesized using the CFR followed by a SAR mixer at a silver nitrate
concentrationofA:0.05mM,B:0.15mM,C:0.25mM,D:0.4mM.Concentrationofsodiumborohydride,0.3mM;
concentrationoftrisodiumcitrate,0.5M;volumetricflowrateratio,1:1(Qout:Qin,NaBH4:AgNO3);molarflowrate
ratio,0.5-4:3:5(AgNO3:NaBH4:Na3citrate);flowrate,2.5ml/min;temperature,22-24°C.
Chapter5
109
Figure 5-11:UV-Vis spectra of silver NPs synthesized using the CFR followed by a SARmixer at a silver nitrate
concentrations ranging between 0.05 and 0.4mM. Inset: Peak absorbance (black squares) and peakwavelength
(reddiamonds)ofsilverNPsvs.silvernitrateconcentrationofsynthesis.Concentrationofsodiumborohydride,0.3
mM;concentrationoftrisodiumcitrate,0.5M;volumetricflowrateratio,1:1(Qout:Qin,NaBH4:AgNO3);molarflow
rateratio,0.5-4:3:5(AgNO3:NaBH4:Na3citrate);flowrate,2.5ml/min;temperature,22-24°C.
Figure5-12:Diametervs.silvernitrateconcentrationfortheCFRoperatedfollowedwithaSAR(blacksquares)and
withoutaSAR(reddiamonds).Concentrationofsodiumborohydride,0.3mM;concentrationoftrisodiumcitrate,
0.5 M; volumetric flow rate ratio, 1: 1 (Qout:Qin, NaBH4: AgNO3); molar flow rate ratio, 0.5-4:3:5
(AgNO3:NaBH4:Na3citrate);flowrate,2.5ml/min;temperature,22-24°C.
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
110
5.2.4 SilvernanoparticlesynthesisusingacoaxialflowdeviceoperatedathighRe
Intheprevioussection(5.2.3),wehaveobserveddifferencesinNPsynthesiswhenmixingconditionsare
altereddownstreamoftheCFR.InthissectionweinvestigatetheeffectofmixingconditionsintheCFR
itself.ThisisachievedbyusingaCFRwithlowI.Dandhighflowrateoftheinnertube,toincreasetheRe.
Even though turbulent conditions are not achieved,mixing is enhanced by recirculation vortices that
appearundertheseconditions.188TheCFRhadaninnertubeI.D.of0.798mmandanoutertubeI.D.of
2mm.ThedistancefromtheinnertubeoutlettotheoutletoftheCFRwas130mm.Theconcentrations
of silver nitrate, trisodium citrate and sodium borohydride were varied so that the concentration
(assumingfullmixingofthestreams)foreachflowratetestedwasfixedat0.1mM,0.5mMand0.3mM
respectively. Silver nitrate and trisodium citrate were pumped through the inner tube and sodium
borohydridewaspumpedthrough theouter tube.Thevolumetric flowrate ratiowasvariedbetween
1:50and1:200 (Qout:Qin) and the total flow ratewasvariedbetween5.1and20.1ml/min.Themolar
flow rate ratio was fixed at 1:3:5 (AgNO3:NaBH4:Na3citrate). Experiments were carried out at room
temperaturebetween22and24°C.
Figure5-13showsTEMimagesofsilverNPssynthesizedintheCFRwithvortexflow.TheReynolds
number intheinnertubewasintherange132-530andinthemainchannelwas intherange54-212.
Theaveragediameteranddispersitywas5.9±1.5nm,6.2±2.7nm,7.7±3.4nmand6.2±2.2nmfor
5.1,10.1,15.1and20.1ml/minrespectively.Figure5-14showsflowvisualizationoftheCFRataflow
rateof20.1ml/mintodemonstratetheturbulentnatureoftheflow.Flowvisualizationusingdyeand
wateratotherflowratescanbeseeninAppendixD(FigureD3).
Flow rates of 5.1, 10.1 and 20.1ml/min showed similar sizes, although 5.1ml/min had a lower
polydispersity. T-tests showed that p-valueswere < 0.001when comparing 15.1ml/min to anyother
flowrate,and<0.05whencomparing5.1and20.1ml/min.Allotherflowratecomparisonsshowedthe
NPpopulationwasnotstatisticallydifferenttoanydegreeofconfidence.
TheNPsobtainedusingtheCFRoperatedinvortexflowshowedthattherewereminimalchanges
and no observable trends as flow rate was increased. Other systems such as iron oxide NPs and
polystyrene NPswere synthesized successfully with low dispersity using a CFR operated in turbulent
flow,131inthiscaseincreasingmixingefficiencyseemedtohavenosignificanteffectonthesynthesized
NPs in terms of size and polydispersity. This suggests that increasing mixing efficiency doesn’t
necessarilyleadtoincreasedmonodispersitybecauseofaperceivedimprovementinmixingefficiency,
although the smallest and least disperse NPs were obtained at the lowest flow rate 5.1 ml/min.
Observingthehydrodynamicsneartheoutletoftheinnertube,theconcentrationofthedye(thiswould
representsilvernitrateintheNPsynthesis)spreadsacrosstheentirecrosssectionofthechannel.The
steadystateconcentrationofdyenearthechannelwalls inthesectionimmediatelyafter jetemission
fromtheoutlet the inner tube ismostdiluteat the lowest flowrateand thensubsequentlybecomes
moreconcentrateathigher flowrates.Thisareathenbecomesmoreconcentrated insilvernitrateas
theflowrate increases.Thiswouldexplainwhythere isa lowersizeanddispersityat thelowest flow
rate,wheretheratioofsilvernitratetosodiumborohydrideislowerandhencemorestableNPswhich
arelesspronetogrowththroughcoalescencearesynthesized.
Chapter5
111
Figure5-13:TEMimageofsilverNPssynthesizedusingtheCFRwith innertubeI.D.of0.798mmoperatedunder
highRewithatotalflowrateofA:5.1ml/min,B:10.1ml/min,C:15.1ml/min,D:20.1ml/min.Theoutertubeflow
rate was fixed at 0.1 ml/min. Concentration of silver nitrate, 0.1 mM; of sodium borohydride, 0.3 mM;
concentrationof trisodiumcitrate,0.5M (all concentrations statedare thoseafter completemixing);molar flow
rateratio,1:3:5(AgNO3:NaBH4:Na3citrate);temperature,22-24°C
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
112
Figure5-14:FlowvisualizationofCFRwithinnertubeI.D.of0.798mmoperatedunderturbulentflowconditions.
BasicBluedyewaspumpedthroughtheinnertubeat20ml/minandwaterwaspumpedthroughtheoutertubeat
0.1ml/min. Re is 520 in the inner tube and 212 in themain channel. Picture 1 is taken before dye is pumped
throughtheinnertubeup,Pictures2-5asoperationaltimeincreased,Picture6aftersteadystateisreached.
5.2.5 EffectofinnertubeinternaldiameteronsilverNPsynthesisusingtheCFR
TheeffectoftheinnertubeI.D.onthesynthesisofsilverNPsintheCFRwasinvestigatedbyvaryingthe
innertubeI.D.between0.147and0.798mm.TheoutertubeI.D.was2mmandthedistancefromthe
innertubeoutlettotheoutletoftheCFRwas130mm.Theconcentrationofsilvernitratewas0.1mM,
whilethetrisodiumcitrateconcentrationwasfixedat0.5mM.Thesetwocomponentswerepremixed
andpumped through the inner tube.Concentrationof sodiumborohydridewas fixedat0.3mM; this
component was pumped through the outer tube. The volumetric flow rate ratio was fixed at 1:1
(Qout:Qin)andthetotalflowratewasfixedat1ml/min.
Figure5-15showsTEMimagesofsilverNPssynthesizedataninnertubediameterof0.147,0.345,
0.447,0.556,0.701and0.798mmwheretheaveragediameteranddispersityoftheNPsare4.7±1.4
nm, 5.9 ± 2.4 nm, 5.9 ± 2.2 nm, 6.6 ± 3.7 nm, 8.8 ± 2.6 nm and 10.5 ± 4.0 nm. T-tests for the PSDs
showedp-valuesof<0.001whencomparingallPSDsexceptwhencomparing0.345and0.447mm(p-
value=0.91),0.345and0.556mm(p-value=0.0012)and0.447and0.556mm(p-value=0.0012). In
thiscasethereisnostatisticaldifferencebetween0.345and0.447mminnertubeI.D.accordingtothe
p-value.0.345and0.447mmarestatisticallydifferentto0.556mmtoasignificancelevelof0.01.Figure
5-16showscorrespondingUV-Visspectra.Thepeakabsorbanceshowedamaximumat0.447mm(UV-
visdataofrepeatscanbeseeninAppendixD,FigureD4)andthepeakwavelengthremainedrelatively
constantbetween392and394nmforalldiameters.
ThesizeoftheNPsincreasedwithanincreaseintheinnertubeI.D.TheinternaltubeI.D.controls
thethicknessoftheinnerstream,andhencethediffusiondistanceacrossthestream.Byusingsmaller
innertubingI.D.,thediffusiondistanceisreduced.Byreducingthediffusiondistance,moreofthesilver
nitrateintheinnerstreamcanreactwithsodiumborohydride.Asthediffusiondistanceincreaseswith
increasingdiameter,thereisalargeramountofunreactedsilvernitrateattheendofthechannel.When
droplets format theoutlet, this increases the likelihoodof reactionsanduncontrolledgrowthofNPs
occurringinsilvernitraterichregionswithinthedroplets.HencesmallerNPsareobtainedbyreducing
thediffusiondistanceofthesilvernitratestreamwithintheCFR.
Chapter5
113
Figure5-15:TEMimagesofsilverNPssynthesizedusingtheCFRwithinnertubediametersofA:0.142mm,B:0.345
mm, C: 0.447 mm, D: 0.556 mm, E: 0.701 mm, F: 0.798 mm, G: Nanoparticle diameter vs. inner tube internal
diameter.Concentrationofsilvernitrate,0.1mM;concentrationofsodiumborohydride,0.3mM;concentrationof
trisodium citrate, 0.5 M; volumetric flow rate ratio, 1: 1 (Qout:Qin, NaBH4: AgNO3); molar flow rate ratio, 1:3:5
(AgNO3:NaBH4:Na3citrate);flowrate,1ml/min,temperature22-24°C.
Effectofhydrodynamicsandmixingconditionsonthecontinuoussynthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
114
Figure 5-16:UV-Vis spectra of silver NPs synthesized using the CFRwith inner tube diameters ranging between
0.142and0.798mm.Inset:Peakabsorbance(blacksquares)andpeakwavelength(reddiamonds)ofsilverNPsvs.
inner tube internal diameter used in the CFR. Concentration of silver nitrate, 0.1mM; concentration of sodium
borohydride,0.3mM;concentrationof trisodiumcitrate,0.5M;volumetric flowrate ratio,1:1 (Qout:Qin,NaBH4:
AgNO3);molarflowrateratio,1:3:5(AgNO3:NaBH4:Na3citrate);flowrate,1ml/min,temperature22-24°C.
5.3 Conclusions
Silver and gold NPs were synthesized in continuous flow mode in a Coaxial Flow Reactor (CFR) in
conjunctionwithothermicrofluidiccomponentssuchasaContinuousFlowInverter(CFI)asaresidence
tile loopforgoldNPsynthesisoraSARmixer forsilverNPsynthesis.Sizecontrolwasachieved inthe
synthesisofgoldNPsbyvarying the flowrate.TheCFIprovideda laminar region inwhichnucleation
occurredat the interfaceof the twostreamsand theCFIprovidedamixing region (due to secondary
flowcirculation)withanimprovedresidencetime(becauseofflowinversion)forgrowthtooccur.Inthis
waynucleationandgrowthperiodscouldbecontrolled,resultinginanNPsizerangeof17.9-23.9nmfor
aflowraterangeof0.25-3ml/min.Aminimumsizeof17.9nmexistedat80°Cwhenthesynthesiswas
testedintemperaturerangeof60-100°C.Inbatchsyntheses,achangeinsizeisachievedbyalteringthe
concentrationsororderofreagentadditionratherthanthemixingconditions.Usingthisapproach,the
systemoffersahigherprecisionofcontroloverthemasstransfer,asopposedtoabatchreactor, ina
relativelysimplemanner(changingtheflowrateofreagents)toachieveavariationinsizeofgoldNPs.
TheCFRusedinconjunctionwiththeSARmixerproducedadecreasingsilverNPsizefrom5.5to3.4nm
with increasing silver concentration,which is the opposite trend to using the CFR in isolation. This is
becausetheSARmixerprovidedveryefficientmixinginawell-controlledmannertopreventreactionin
silvernitrate rich regions.TheCFRalsoallows someof the reaction tooccurbeforeentering theSAR
Chapter5
115
mixer.ThisreducedthefoulingoccurringintheverysmallandcomplexgeometryoftheSAR(usingthe
SARinisolationresultedinahigheramountoffouling).AvortexflowregimeathigherRenumberwas
testedintheCFRwithsizerangeof5.9-7.7nmforaflowraterangeof5.1-20.1ml/min.Notrendinsize
or dispersity was observed using this type of flow regime, even though the mixing efficiency was
increasedwithincreasingflowrate.Interestingly,thesmallestandleastdisperseNPswereobservedat
thelowestflowrate,whichwasmostlikelybecauseoftheregionclosetotheoutletoftheinnertube.
Thespreadofsilvernitrateacrossthechannelundertheseconditionswasrelativelysmallcomparedto
higherflowratesandNPspronetolessgrowthwouldbeformedinregionsdiluteinsilvernitrateand
concentratedinsodiumborohydride.DecreasingtheinnertubeI.D.intheCFRresultedinadecreasein
NP size from10.5 to4.7nm,due to the reduceddiffusiondistanceof the inner streamresulting ina
fasterconsumptionofsilvernitrateinawell-controlledmasstransferregion(i.e.withintheCFRrather
thanat theoutlet),whichwas supportedbyCFDmodellingof the reactionwithin theCFR.The study
demonstrates the versatility of using microfluidic devices and the potential benefits arising from
manipulatingthemasstransferandhydrodynamicsonNPsizeanddispersitybyusingtheCFRand/or
othercomponentsinconjunctionwiththeCFRforsynthesis.Thislevelofcontrolovermasstransferis
noteasilyachievedusingbatchreactors.
116
6 Masstransfereffectsinthebatchsynthesisofsilverandgoldnanoparticles
TherearemanysyntheseswhichinvestigatehowconcentrationsaffectthesizeoftheNPsandnaturally
different conditions yield different sizes. One factor that is rarely investigated in depth in batch
synthesisstudiesisthemasstransfer,whichisanimportantfactorsinceitalsohasamajorroletoplay
indeterminingatwhatconcentrationsreactionsoccurinlocalizedregionswithinthereactor.
Thebatch reactorsused in the following chapter employ activemixingwhereexternal energy is
appliedtomixthefluidswithin(inthiscaseamagneticstirbarandstirrermotor).Thistypeofmixing
enhancesadvectioninthefluid,leadingtomoreefficientmixing.Themicrofluidicdevicesdonotrelyon
external energy being applied, but rather their small dimensions and manipulations of the channel
geometrytoenhancemixing.Asdiscussedintheintroduction(Chapter1),masstransporthastwomain
mechanismswhicharediffusionandadvection.Microfluidicdevices relymainlyondiffusion formass
transport,incontrastthebatchreactorsusedinthefollowingchapteralsoemployadvectionasamajor
componenttotheoverallmasstransport.
SilverNPsarecommonlysynthesizedusingsodiumborohydrideasareducingagentwithinabatch
reactorvessel(chapter2).5,64-70Ingeneral,alteringthemolarratiosofreducingagenttoprecursorwas
themainmethod toachieve controlover the sizeof the silverNPs. Thishighlights the importanceof
concentration intheresultantNPsize.SimilarlyforgoldNPs,theTurkevichmethodisaverycommon
method employed. This involves the reduction of tetrachloroauric acid via the citratemolecule at an
elevated temperature. Many studies investigate this method for synthesis of gold NPs using batch
reactors (chapter 2).88-98, 100-106 The batch syntheses studies on the Turkevich method investigate
changing theorderof reagentaddition,but there isa lackof studieswhich focusspecificallyonmass
transferandhowitaffectstheresultantNPsizeanddispersity.
ThisstudytakesthesetwopopularmethodsforthesynthesisofsilverNPsandgoldNPsinbatch
vessels, and investigates the role mass transfer plays in the resultant size and dispersity of the
respectiveNPsystem.ThemixingtimeischaracterizedusingtheVillermaux-Dushmanreactionscheme
and various mixing configurations and order of addition of the reagents is investigated to draw
conclusionsontheeffectofmasstransfer.Forallofthemixingconfigurations,thefinalconcentration
aftermixingresultedinthesameconcentrationofreducingagentandprecursor.
6.1 Methodology
6.1.1 Chemicals
Silver nitrate (AgNO3, 0.01 M stock solution), trisodium citrate (HOC(COONa)(CH2COONa)2·2H2O,
powderform,sodiumborohydridesolution(NaBH4,~12wt%in14MNaOHstocksolution)andgold(III)
chloridehydrate(HAuCl4.xH2O,powderform),potassiumiodide(KI,99%)potassiumiodate(KIO3,98%)
andboricacid(H3BO3,99.5%)wereobtainedfromSigma.Sulfuricacid(H2SO4,3Mstocksolution)was
Chapter6
117
obtained from Alfa Aesar. All chemicals were used without further purification and solutions were
preparedwithultrapurewater(resistivity15.0MΩ·cm).
6.1.2 ExperimentalSetup
Syringepumps (Pump11EliteOEMmodule,Harvard) connected toa0.5mm internaldiameter (I.D.)
316LstainlesssteeltubeforsilverNPsanda0.5mmI.D.PTFEtubeforgoldNPswereusedtodeliver
oneofthereagentstothebatchvessel.Thebatchvesselwasa50mlsmallneckglassErlenmeyerflask
(Fisher Scientific). A digital hotplate stirrerwith temperature control (Stuart)was used togetherwith
PTFEstirbars(15mmx1.5mm,VWRinternational)tomixthereagentswithinthevessel.Aglycerine
bathwasusedtogetherwiththetemperaturecontrollertoheatthereagentswithinthebatchvesselin
thecaseofthegoldNPsynthesis.Acapwasusedtocoverthebatchvesselandthetubewasinserted
throughthecapforadditionofreagents.Thelengthofthetubecouldbeadjustedtoallowadditionof
reagents fromabove the solutionornear the tipof themagnetic stirbar.Thiswasdone toallow for
variation inmixingconditions.Figure6-1showsaschematicof theexperimental setups for silverand
goldNPsynthesis.
Figure6-1:SchematicofexperimentalsetupforsilverandgoldNPsynthesisconsistingofanErlenmeyerflaskwith
precursororreducingagentstirredviaamagneticstirrerwithasyringepumpreagentviatubinga)silverNPsystem
usingastainlesssteel0.5mmI.D.tubingandb)goldNPsystemusingaPTFE0.5mmI.D.tubingwithaglycerine
bathandtemperatureprobefortemperaturecontrolat80°C.Stirrerspeedwassetat500rpminbothcases.
Masstransfereffectsinthebatchsynthesisofsilverandgoldnanoparticles
118
6.1.3 Nanoparticlesynthesis
InatypicalsilverNPsynthesis,20mlofreagentwereplacedwithinthebatchvesselandstirredat500
rpm.Tothis,0.6mloftheotherreagentwasaddedviathesyringepumpandtubeateither:0.5ml/min
fromabove thesolution indropletsor,50ml/minwith the tube inserted into thesolutionwithin the
batchvesselnearthestirbartip.Thereagentswereamixtureofsilvernitrateandtrisodiumcitrate,and
thesodiumborohydridesolution.Theorderofmixingwastestedbyswitchingwhichreagentwasplaced
inthebatchvessel,either20mlofsilvernitrateandtrisodiumcitrateor20mlofsodiumborohydride
(sodiumborohydridewasstored in14Msodiumhydroxide leadingtoNaOHconcentrationbeing3.21
timeshigherthanthesodiumborohydrideconcentrationinallcases).Similarly,inthegoldNPsynthesis
20mlof reagentwasplaced in thebatchvessel and0.6mlof reagentwasadded to this in a similar
mannertothesilverNPsynthesis.20mloftetrachlorauricacidor20mloftrisodiumcitratewereplaced
inthebatchvessel.Thissolutionwasplacedintheglycerinebathwhichwasheatedto80°C.Totestthe
lengthoftimerequiredtobring20mlofsolutionto80°C,temperaturewasmeasuredovertimeafter
inserting solutionat room temperature into theglycerinebathand itwas found that after7min the
temperature reached a steady state of 80°C. Hence the solution in the experimentswas heated and
stirredfor15minbeforeadditionof0.6mlofreagentfromthesyringepump.
6.1.4 Characterizationofnanoparticles
ForUV-Visanalysis,NPswereanalyzedusingaUV-Visspectrometer(USB2000+SpectrometerandDT-
Mini-2-GSlightsource,OceanOptics).SilverNPswereanalyzedwithinanhourofsynthesis(thesignalof
thesampleswerestableinthiswindowoftime).GoldNPswereanalyzedovertimeandwereanalyzed
oncethesignalwasstable(usually1-2days).NPsamplesweredilutedwithadditionalultrapurewaterto
bringtheabsorbanceintoasuitablerange(i.e.obeyingtheBeer-Lambertlawandavoidingsaturationof
the lightdetector) ifnecessary.Transmissionelectronmicroscope imageswerecapturedusinga JEOL
1200 EX ii microscope with a 120 kV acceleration voltage. Carbon coated copper TEM grids were
preparedwithin anhourof synthesis for silverNPs andwhen the samplewas stable for goldNPsby
pipettinga5µlsampleontothegridandallowingittodryatroomtemperature.Carewastakentonot
covertheentireTEMgridwiththedroplet,sothattheaggregatedringaroundtheedgeofthedroplet
couldeasilybeidentifiedasanartefactofsamplepreparation.Particlesizedistributions(insetsforeach
TEM image presented) have the following nomenclature: d is average diameter, δd is the standard
deviation of the NP distribution and n is the number of particles counted to obtain the particle size
distribution. Differential centrifugal sedimentation analysis (CPS disc centrifuge UHR, Analytik) was
carriedoutongoldNPswhenthesampleswerestable.PEBBLESsoftwarewasusedforthecountingand
sizingofTEMimagesofthesynthesisedsilverNPsandImageJsoftwarewasusedforthecountingand
sizingofTEMimagesofthesynthesisedgoldNPs.
Chapter6
119
6.2 Resultsanddiscussion
6.2.1 MixingCharacterization
1mlof0.07Msulfuricacidwasaddedto20mlofabuffersolutionconsistingof0.1818Morthoboric
acid,0.0909Msodiumhydroxide,11.67mMpotassiumiodideand2.33mMpotassiumiodate.Stirrer
speedwas set to 500 rpm and the twomixing configurations used in the NP synthesis were tested.
Thesewere:acidwasinjectedfromthecentreabovethebuffersolutionornearthestirbartip.Figure
6-2showsaschematic illustratingthetwoconfigurations.Figure6-3showsthecomputedsegregation
indexvs.mixingtimefortheconcentrationsetoftheexperimentsusingtheIEMmixingmodel(details
can be found in Appendix A). Eachmixing configuration was repeated three times and the triiodide
concentration was measured (extinction coefficient was found to be 23209 l/mol.cm, details on the
procedure used to find this can be found in the Appendix A). Themixing time averaged over three
repeatexperimentsforthe injectionfromabovethesolutionwas209mswhilethe injectionnearthe
stirbartipwas46ms.Areductioninmixingtimewasexpectedsincethereisamuchhighershearrate
near the stir bar tip, resulting in improvedmixing efficiency. Since both experiments used the same
concentrationset,injectionnearthestirbartipwasapproximately4timesmoreefficienttheninjection
fromabovethebuffersolution.
Figure6-2:Schematicrepresentationofinjectionpointswithinthebatchvesseltovarymixingefficiencya)injection
isaddedfromabovethesolutioninthebatchvesselandb)injectionisaddedclosetothestirbartip.
Masstransfereffectsinthebatchsynthesisofsilverandgoldnanoparticles
120
Figure6-3:Segregationindexvs.mixingtimeestimatedusingtheIEMmodelfortheVillermaux-Dushmannreaction
scheme.ConcentrationsofbuffersolutionwereKI:0.0117M,KIO3:0.0023M,NaOH:0.0909M,H3BO3:0.1818M.
ConcentrationofH2SO4was:0.07M.
6.2.2 Silvernanoparticlesynthesis
SilverNPsweresynthesizedinthebatchvesselbyreductionofsilvernitrateviasodiumborohydridein
the presence of trisodium citrate. The effect of changing the mixing configuration and the effect of
changing theorderof reagentadditionon thesizeanddispersityof theNPswas investigated.Mixing
configurations were those described in the previous section i.e. injection of reagent from above the
solutionorinjectionnearthestirbartip,andtheorderofreagentadditionwasachievedbyswapping
thereagentthatstartedinthebatchvesseli.e.asilvernitrateandtrisodiumcitratemixtureorsodium
borohydride. The general procedurewas to inject 0.6ml of reagent via tubing into 20ml of reagent
whichwasbeingstirredinthebatchvessel.FourconditionsweretestedandwillbedescribedasA-Ag-R-
S,B-Ag-R-F,C-Ag-P-SandD-Ag-P-F.Theorderofreagentaddition isdenotedbythe lettersRandP,R
representstheadditionof0.6mlofreducingagenttothebatchvesselandPrepresentstheadditionof
0.6mlofprecursortothebatchvessel.ThemixingconfigurationsaredenotedbythelettersSandF,S
denotes the additionof 0.6ml of reagent via droplets at a rateof 0.5ml/min fromabove the20ml
mixturewhileFdenotestheadditionof0.6mlofreagentnearthestirbartipatarateof50ml/min(for
example, the condition A-Ag-R-S describes the addition of 0.6 ml of reducing agent at a rate of 0.5
ml/minto20mlofprecursor).Table6-1summarizestheconcentrationsforeachmixingcondition(3to
1molarratioofborohydridetosilvernitratewasusedinallcases).
Chapter6
121
Table6-1:SummaryofconcentrationsforeachmixingconditionusedinthesilverNPsynthesis
Condition ReducingAgent–NaBH4 Precursor–AgNO3/Na3citrate
A-Ag-R-S 63.3mM 0.38/0.38mM
B-Ag-R-F 63.3mM 0.38/0.38mM
C-Ag-P-S 1.36mM 9.1/9.1mM
D-Ag-P-F 1.36mM 9.1/9.1mM
Figure6-4showsTEMimagesofthesynthesizedsilverNPswithanaveragediameterof11.1±3.0nm,
6.7±1.7nm,9.5±1.4nmand11.5±2.4nmforconditionsA-Ag-R-S,B-Ag-R-F,C-Ag-P-SandD-Ag-P-F
respectively. The average peak absorbance and average wavelength obtained from 3 repeat
experimentsobtainedbyanalysing thesilverNPsusingUV-vis spectroscopycanbeseen inFigure6-5
(thepercentageofdeviation inpeakabsorbance foreach conditionwas:A-Ag-R-S=3.2%,B-Ag-R-F=
2.1%, C-Ag-P-S = 2.2% and D-Ag-P-F = 2.7%). The peak wavelength had increased deviation for
conditionsA andDon repeat experiments.Overall therewas little variation in the valueof thepeak
wavelengthacrossalltheconditionstested.SincetheNPsare lessthanaround11nm, it isconsistent
with thebehaviour of little variation in peakwavelength forNPs below10nm.139 ConditionB-Ag-R-F
produced the smallest NPs followed by condition C-Ag-P-S, while conditions A-Ag-R-S and D-Ag-P-F
produced NPs of comparable size. The smallest NPs were produced when a small volume of
concentratedsodiumborohydridewasaddedquicklyintothedilutesilvernitrateandtrisodiumcitrate
mixture, orwhendrops of a concentratedmixture of silver nitrate and trisodium citratewere added
slowlytoasodiumborohydridesolution(conditionBandCrespectively).Conversely, largerNPswere
synthesizedwhendropsofconcentratedsodiumborohydridewereaddedslowlytoadilutemixtureof
silvernitrateandtrisodiumcitrateorwhenaconcentratedmixtureofsilvernitrateandtrisodiumcitrate
wereaddedquicklyintoadilutevolumeofsodiumborohydride.Inthissynthesis,trisodiumcitrateisthe
ligandandsinceitismixedwithsilvernitratepriortoadditionofsodiumborohydride,itcanbeassumed
that the ligand is available in the vicinity where silver nitrate is reduced by sodium borohydride.
However it is known that sodiumborohydride also behaves as a stabilizing agent,where the surface
charge of the particles is affected by borohydride ion adsorption.136 The difference in sizes obtained
using the different conditions is due to the availability of sodium borohydride in the local region
surroundingareaswheresilvernitratehasbeenreducedtosilvermetal.Thesilvermetalsubsequently
produces small silver clusters which are stabilized by sodium borohydride. Silver NPs catalyse the
hydrolysis of sodium borohydride which produces hydrogen and normally this catalysis leads to a
switchingpointwheretheNPssuddenlyincreaseinsizeduetoaggregation,68whichoccursbecausethe
stabilizing presence of borohydride ions is depleted through the hydrolysis reaction which provides
evidencefortheirstabilizingeffect.NoswitchingpointisobservedinthisstudyduetoanelevatedpH
becauseofthepresenceofsodiumhydroxide,suppressingthehydrolysisreaction.137Anotherimportant
point to note is that sodiumborohydride can reducemore than its own stoichiometric equivalent of
silver ions since it can provide multiple electrons in the reduction reaction.133-135 Therefore a small
amountofsodiumborohydrideca3enreducemanysilverions(whichsubsequentlyformsilverNPs),but
Masstransfereffectsinthebatchsynthesisofsilverandgoldnanoparticles
122
there may be a deficient amount of sodium borohydride to provide stabilization of NPs resulting in
growthtoalargersize.Thiswouldoccurthroughcoalescenceofsmallersilverclustersintolargersizes,
thereducedsurfacechargeresultinginthepossibilityofclustersapproachingeachothercloseenough
for attractive forces tobring them together. Coalescencehasbeen identified inprevious studies as a
maingrowthmechanisminthesilverNPformationprocess.68,82,139
Consideringthepropertiesandroleofthesodiumborohydrideinthereaction,aswellasthemass
transportmechanismswithinthebatchreactor,canhelptoexplaintheresultsobtained.Thesizeofthe
NPs obtained under conditions A-Ag-R-S and D-Ag-P-F can explained as follows: the conditions
potentiallycreatemixingconditions inwhichsilvernucleiareexposedtodeficientamountsofsodium
borohydridewithanabundanceofprecursorleadingtotheirgrowthtoalargersize.ConditionA-Ag-R-S
has concentrated sodium borohydride which is added slowly drop by drop. Since there is enhanced
advectionwithinthebatchvesselbecauseofthestirringaction,eachdropwillbemixedwiththebulkof
thesilvernitrate/trisodiumcitratesolutionbeforesubsequentdropsareadded.Thefirstfewdropscan
reduce themuch largermassof silver nitrate in thebatch vesselwhich createsNPsprone to growth
becauseofdeficiencyinsodiumborohydride.ConditionD-Ag-P-Fhasconcentratedsilvernitrateadded
quickly to dilute sodium borohydride. Since there is a sudden influx of a largemass of silver nitrate
becauseaddition isquick (0.6mlof fluid takes0.72 seconds toaddata flow rateof50ml/min), the
reactionofsilvernitratecanoccurinlocalizedregionsinthemicroscalewhereitishighlyconcentrated.
Thiswouldagain createNPsprone togrowth since the reactionand subsequentnucleationoccurs in
areaswherethesilvernitratetosodiumborohydrideratioishigh,resultinginalargerfinalsize.Inthis
case,ratherthantheenhancedadvectioncausinggrowthoftheNPs(asisthecaseforconditionA-Ag-R-
S),theslowdiffusiononthemicroscaleaftertheinfluxofasmallconcentratedvolumeofsilvernitrate
wouldcausereactionsinareasthatarerichinsilvernitrate.Thisleadstothepossibilityoflargenuclei
forming at the very beginning of the reaction because of decreased stability (sodium borohydride
deficiency) and would result in fewer nuclei which would subsequently grow into larger NPs. The
smallest NPs are obtained under the condition B-Ag-R-F, when a concentrated volume of sodium
borohydride isaddedquicklytoadilutevolumeofsilvernitrate.Comparedtothecasewheresodium
borohydrideisaddedslowly,inthiscasealargemassofsodiumborohydridebecomesavailablequickly
to the dilute volume of silver nitrate. This would create a condition in which there is initially an
abundanceofsodiumborohydrideinlocalizedregionsinthemicroscalebeforetheenhancedadvection
causescompletemixinginthebulkofthefluidwithinthevessel,resultinginsilverNPswhicharevery
stableandhencenotpronetogrowth.Asthemixingproceedsandcompletesonthemacroscale, the
sodiumborohydrideconcentrationwouldreduceduetomixingwithalargerdilutevolumeandbecause
of reaction that has occurred during the mixing process. The reaction can be visibly seen as a dark
brownplumeonadditionofsodiumborohydride immediatelyafter injection,afterwhichthesolution
transitionsintoabrightyellowcolouroncethemixturehasbeenhomogenizedthroughmixing.Thedark
brownplumecouldbe indicativeofaggregatedsilverNPsandthe transition toabrightyellowcolour
(indicativeofnon-aggregatedsilverNPs)maybebecauseofrapidacquisitionofstabilisingborohydride
causingtheaggregatedNPstorepeleachotherandseparate.Asthesodiumborohydrideconcentration
Chapter6
123
isreduced,reactionmayoccurinsomeregionswherethesodiumborohydridetosilvernitrateratiois
lowerandhencethepossibilityofsomelargerNPsmaybeproduced.Infact,thisiswhatisseeninthe
particlesizedistributionforconditionB-Ag-R-F.ThereisarelativelylargeamountofsmallNPs(around6
nm)with relatively small amountsof largerNPs (around10nm). Finally, conditionC-Ag-P-Sproduced
the most monodispersed NPs and these were of an intermediate size. Concentrated drops of silver
nitratewereadded toadilutevolumeof sodiumborohydride. In this case,each silvernitratedrop is
exposedtoamuchhighermassofsodiumborohydridesoisthereforeexpectedtomakestableNPsthat
arenotpronetogrowth.However,sincethemicromixingtimeisslowerusingthismixingconfiguration,
it is expected that some reactionwill take place as the concentrated silver nitrate drop ismixed on
addition to the dilute volume of sodium borohydride. This explains the intermediate size obtained,
wherethelocalizedsilvernitrateconcentrationishighbutissubsequentlyexposedtoanabundanceof
sodiumborohydride.Thecombinationofdiffusionandadvectionandhowitcanoffercontroloverthe
NPsizeisdemonstratedwellintheseexperiments.Becausediffusionisslow,largernucleicouldformin
thesilvernitraterichdropasitentersthebulkfluid(wherereactionoccursinsilvernitraterichregion
withinthefluidintheimmediatevicinityofthedrop)butthisisquicklysuppressedasadvectionaidsin
mixingthesilvernitratedroplet intothebulksodiumborohydridefluidtostabilise it. Inessence,each
dropof silvernitrate reacts to formsilverNPswhicharequickly stabilizedas theyaremixed into the
sodiumborohydridesolution,andfurtherdropsthenproduceNPswhichdonot interactwithexisting
NPsinthesolutionbecauseoftheirstabilitywhichexplainsthelowerdispersityoftheresultantNPs.It
is theconfigurationwhichgives themostuniformsodiumborohydrideconcentrationexposure to the
silverNPs,andpointstothesodiumborohydridebeingcrucialindeterminingtheNPsizeanddispersity
in this synthesis. Shirtcliffe et al. also compared different orders of reagent addition and found that
adding silver nitrate to sodium borohydride produced the narrowest and most reproducible UV-VIS
spectra, pointing to lower dispersity.65 Song et al. also noted that increasing sodium borohydride
resulted in decreasing aggregation, again highlighting the importance of sodium borohydride to the
stabilityofthesilverNPs.67
Masstransfereffectsinthebatchsynthesisofsilverandgoldnanoparticles
124
Figure 6-4: TEM images of silver NPs synthesized in a batch vessel under various conditions. A) 0.6 ml sodium
borohydride [63.3mM] added via drops to 20ml of amixture of silver nitrate [0.38mM] and trisodium citrate
[0.383mM]ataflowrateof0.5ml/min,B)0.6mlsodiumborohydride[63.3mM]addednearthestirbartipto20
mlofamixtureofsilvernitrate[0.38mM]andtrisodiumcitrate[0.383mM]ataflowrateof50ml/min,C)0.6mlof
amixtureofsilvernitrate[9.1mM]andtrisodiumcitrate[9.1mM]addedviadropsto20mlsodiumborohydride
[1.36mM]ataflowrateof0.5ml/minandD)0.6mlofamixtureofsilvernitrate[9.1mM]andtrisodiumcitrate
[9.1 mM] added near the stir bar tip to 20 ml sodium borohydride [1.36 mM] at a flow rate of 50 ml/min.
Experimentswereconductedatroomtemperatureandstirringspeedwassetat500rpm.
Chapter6
125
Figure 6-5:Averagepeak absorbance and averagepeakwavelengthobtainedusingUV-vis spectroscopyof silver
NPssynthesizedinbatchvesselundervariousconditions.A)0.6mlsodiumborohydride[63.3mM]addedviadrops
to20mlofamixtureofsilvernitrate[0.38mM]andtrisodiumcitrate[0.383mM]ataflowrateof0.5ml/min,B)
0.6mlsodiumborohydride[63.3mM]addednearthestirbartipto20mlofamixtureofsilvernitrate[0.38mM]
andtrisodiumcitrate[0.383mM]ataflowrateof50ml/min,C)0.6mlofamixtureofsilvernitrate[9.1mM]and
trisodiumcitrate[22.7mM]addedviadropsto20mlsodiumborohydride[1.36mM]ataflowrateof0.5ml/min
andD)0.6mlofamixtureofsilvernitrate[9.1mM]andtrisodiumcitrate[22.7mM]addednearthestirbartipto
20 ml sodium borohydride [1.36 mM] at a flow rate of 0.5 ml/min. Experiments were conducted at room
temperature and stirring speed was set at 500 rpm. Bars represent standard deviation of absorbance and
wavelengthfor3repeatedexperiments.
6.2.3 Goldnanoparticlesynthesis
GoldNPsweresynthesizedinthebatchvesselbyreductionoftetrachloroauricacidviatrisodiumcitrate
atatemperatureof80°C.Theeffectofchangingthemixingconfigurationandtheeffectofchangingthe
order of reagent addition on the size and dispersity of the NPs was investigated. The same four
conditionstothoseusedforsilverNPsynthesisweretestedandwillbedescribedasconditionsA-Au-R-
S,B-Au-R-F,C-Au-P-SandD-Au-P-F.Table6-2summarizestheconcentrationsforeachmixingcondition
(3to1ratioofcitratetotetrachloroauricwasusedinallcases).
Masstransfereffectsinthebatchsynthesisofsilverandgoldnanoparticles
126
Table6-2:SummaryofconcentrationsforeachmixingconditionusedinthegoldNPsynthesis
Condition ReducingAgent–Na3citrate Precursor–HauCl4
A-Au-R-S 92.7mM 0.556mM
B-Au-R-F 92.7mM 0.556mM
C-Au-P-S 2.78mM 18.5mM
D-Au-P-F 2.78mM 18.5mM
Figure6-6showsTEMimagesofthesynthesizedgoldNPswithanaveragediameterof18.0±4.8nm,
13.1±2.2nm,16.3±3.3nmand16.2±2.1nmforconditionsA-Au-R-S,B-Au-R-F,C-Au-P-SandD-Au-P-F
respectively. The average peak absorbance and average wavelength obtained from 3 repeat
experimentsobtainedbyanalyzingthegoldNPsusingUV-visspectroscopycanbeseeninFigure6-7(the
percentageofdeviationinpeakabsorbanceforeachconditionwas:A-Au-R-S=0.8%,B-Au-R-F=1.6%,
C-Au-P-S=0.5%andD-Au-P-F=0.2%).ThepeakwavelengthofNPssynthesisedusingconditionB-Au-R-F
was slightly lower than the other conditions in line with the average size of the NPs being smaller.
Diameterwasfoundtobe14.3±3.3nm,13.8±1.7nm,14.2±2.8nmand14.1±2.1nmbasedona
numberpercentageobtainedfromDCSmeasurementsforconditionsA-Au-R-S,B-Au-R-F,C-Au-P-Sand
D-Au-P-Frespectively.ThecurvesanddiametervsconditioncanbefoundinFigure6-8.
The largestandmostpolydisperseNPswereobservedusingconditionA-Au-R-Sandthesmallest
NPs were observed using conditions B-Au-R-F. Conditions C-Au-P-S and D-Au-P-F produced NPs of a
similarsizethoughD-Au-P-FproducedslightlymoremonodispersedNPs.ThereactionsintheTurkevich
method are complex because of the many different species which can coexist in the solution as it
proceeds. The precursor can exist in increasingly hydroxylated form fromAuCl4- toAuOH4
- ,95with the
hydroxylatedformbeingformedatelevatedpH.Accordingtothe literature, increasinglyhydroxylated
formsresultinadecreaseinreactivity.104Thiswouldresultinalowernucleationrateandhencelarger
NPs.Thecitratespeciesalsochanges itsspeciationdependentonpH.At lowerpH itexists in itsmost
protonated form of H3Cit with increasing deprotonation to Cit3- as pH is elevated while the HCit2-
speciesissuggestedtobethemostlikelyformresponsibleforreductionofthegoldprecursor.106Hence,
themixingofthesetworeagentsbecomesimportantbecausethecombinationoftheacidicprecursor
withthebufferreducingagentresultsinanevolutionofpHasmixingproceeds.Ideally,toachievesmall
monodispersedNPs,thenucleationperiodshouldbemaintainedtoproducemanynuclei(resultingina
small size) and there should be a good separation between nucleation and growth (resulting in
monodispersity). Nucleation occurs when there is high reaction rate, which depends on less
hydroxylatedformsofprecursorandmonoprotonatedcitratebeingpresent.Aseparationofnucleation
andgrowthisachievedbyatransitionfromhighreactionratetoalowerreactionrate.Thisoccursinthe
Turkevichmethodbecausethebuffercapacityofthecitratesolutionneutralizestheacidicprecursoras
mixingproceedswhilethepHincreasesastheprecursorreacts.Consideringtherolesofprecursorand
reducing agent, the results obtained from each condition can be explained. Condition A-Au-R-S
produced the largest and most disperse NPs. Concentrated citrate is added drop wise to the dilute
precursor.RelativelylargeNPsareproducedinthiscasebecausecitrateisaddedslowly(0.6mltakes72
Chapter6
127
s toaddata rateof0.5ml/min), resulting in the reduction reactionoccurringata relatively lowrate
sincethecitrateconcentrationwillbeperpetuallylow.Thisisthecasebecauseoftheenhancementsin
mass transport throughadvectionbecauseof the stirringmotion,whichhastens the incorporationof
citrate into the bulk fluid which causes dilution of the concentration. A low reduction reaction rate
results ina lowernucleation rateandhence largerNPs.Thepolydispersenatureof the samplearises
from theevolving citrate concentrationas the rateof additionwasquite slow. Initially there is a low
citrate to gold ratio, resulting in larger and less spherical shapedNPs.Over time this ratio decreases
resultinginsmallerandmoresphericalNPs.ConditionB-Au-R-FproducedthesmallestNPsbecausethe
citratewasaddedatahigherflowratenearthestirbartip.Thisresultedintheprecursorbeingexposed
quicklytoalargeramountofreducingagentincomparisontoconditionA-Au-R-S,resultinginahigher
reaction rate and subsequent amounts of nucleation leading to smaller NPs. In this case, the
concentratednatureofthecitrateleadstoahighreductionrateinthelocalvicinity,resultinginsmaller
and more numerous nuclei. As mixing proceeds through the aid of advection, the increased
concentration of nuclei would ultimately grow to a smaller size assuming the remaining unreacted
precursor is spreadequallyacrosseachnucleus.ConditionC-Au-P-SandconditionD-Au-P-Fproduced
NPs of a similar size. The addition of 0.6 ml to a larger volume of 20 ml wouldn’t reduce the
concentration of the larger volume significantly. In this case, the length of time required to add the
precursorisnotsignificantsinceit isallexposedtoapproximatelythesameconcentrationofreducing
agentresulting insimilarreactionratesandnucleationrates,which iswhyconditionsC-Au-P-SandD-
Au-P-FproduceasimilarsizeofNP.Sivaramanetal.notethatthereisnosignificantreductioninsizeof
goldNPsusingtheinversemethodforcitratetogoldmolarratios<5,102andweobservenosignificant
reductionaccording to this studyatacitrate togoldmolar ratioof5.Thequickadditionofcitrate to
tetrachloroauricaciddidyieldsmallerNPs,butthemanner inwhichreagentsareaddedhasnotbeen
thespecificfocusofotherstudiesofgoldNPsynthesesinbatchreactors.Anotherpointtonoteisthat
thefasteradditionratenearthestirbartipingeneralhadalowerdispersitythantheslowadditionvia
drops.Consideringthatthecitratebehavesasareducingagentandstabilizingagenttheconditionscan
becomparedwithrespecttothisreductionindispersity.Incomparingtheconditions,thisdifferencein
dispersitycanbeexplainedwiththeevolutionofconcentrationswithinthestirredfluid.Fastadditionof
citratecreatesburstnucleationinconditionB-Au-R-F,afterwhichthesolutionpHincreasesresultingin
a transition to growth. Slow addition of citrate in condition A-Au-R-S results in nucleation initially,
subsequentadditionofcitratetheninteractsnotonlywithgoldprecursorbutalsogoldnuclei.Thisgives
the possibility of both nucleation and growth occurring at the same time. In fact, this results in
somewhatofanoverlapbetweengrowthandnucleation.Inessence,thereisasharperseparationusing
conditionB-Au-R-F than conditionA-Au-R-S, resulting in adecreaseofdispersitywith fast additionof
citrate.Similarly,withconditionC-Au-P-SandD-Au-P-F,addingtetrachlorauricacidslowlyresultsinthe
possibilityofanoverlapofnucleationandgrowth,whereasthefastadditionwouldresult inasharper
separationofnucleationandgrowthresultinginlowerdispersity.
Masstransfereffectsinthebatchsynthesisofsilverandgoldnanoparticles
128
Figure 6-6:TEM images of goldNPs synthesized in a batch vessel under various conditions. A) 0.6ml trisodium
citrate[92.7mM]addedviadropsto20mloftetrachloroauricacid[0.556mM]ataflowrateof0.5ml/min,B)0.6
mltrisodiumcitrate[92.7mM]addednearthestirbartipto20mloftetrachloroauricacid[0.556mM]ataflow
rateof50ml/min,C)0.6mlof tetrachloroauricacid [18.5mM]addedviadrops to20ml trisodiumcitrate [2.78
mM]ataflowrateof0.5ml/minandD)0.6mloftetrachloroauricacid[18.5mM]addednearthestirbartipto20
mltrisodiumcitrate[2.78mM]ataflowrateof50ml/min.Experimentswereconductedat80°Candstirringspeed
wassetat500rpm.
Chapter6
129
Figure6-7:PeakabsorbanceandpeakwavelengthobtainedusingUV-visspectroscopyofgoldNPssynthesized in
batch vessel under various conditions. A - Au) 0.6ml trisodium citrate [92.7mM] added via drops to 20ml of
tetrachloroauricacid[0.556mM]ataflowrateof0.5ml/min,B-Au)0.6mltrisodiumcitrate[92.7mM]addednear
the stir bar tip to 20 ml of tetrachloroauric acid [0.556 mM] at a flow rate of 50 ml/min, C - Au) 0.6 ml of
tetrachloroauricacid[18.5mM]addedviadropsto20mltrisodiumcitrate[2.78mM]ataflowrateof0.5ml/min
andD-Au)0.6mloftetrachloroauricacid[18.5mM]addednearthestirbartipto20mltrisodiumcitrate[2.78
mM]ataflowrateof0.5ml/min.Experimentswereconductedat80°Candstirringspeedwassetat500rpm.Bars
representstandarddeviationofabsorbanceandwavelengthfor3repeatedexperiments.
Masstransfereffectsinthebatchsynthesisofsilverandgoldnanoparticles
130
Figure6-8:NormalizednumberpercentagevsdiameterobtainedusingDCSofgoldNPssynthesizedinbatchvessel
undervariousconditions.A-Au)0.6mltrisodiumcitrate[92.7mM]addedviadropsto20mloftetrachloroauric
acid[0.556mM]ataflowrateof0.5ml/min,B-Au)0.6mltrisodiumcitrate[92.7mM]addednearthestirbartip
to20mlof tetrachloroauricacid [0.556mM]ata flowrateof50ml/min,C -Au)0.6mlof tetrachloroauricacid
[18.5mM]addedviadropsto20mltrisodiumcitrate[2.78mM]ataflowrateof0.5ml/minandD-Au)0.6mlof
tetrachloroauricacid[18.5mM]addednearthestirbartipto20mltrisodiumcitrate[2.78mM]ataflowrateof
0.5 ml/min. Experiments were conducted at 80°C and stirring speed was set at 500 rpm. Inset is an average
diameterforeachconditionobtainedfrom3repeatexperiments.
6.3 Conclusions
Mixingefficiency inthebatchvesselwascharacterizedusingtheVillermaux-Dushmanreactionsystem
fortwodifferentmixingconfigurations:injectionofreagentfromabovethestirredsolutioninthebatch
vesselandinjectionofreagentnearthestirbartipofthestirredsolution.Itwasfoundthatthemixing
timewas209mswheninjectionwasfromaboveand46mswheninjectionwasnearthestirbartip.
SilverNPsweresynthesizedbyreducingsodiumborohydrideinthepresenceoftrisodiumcitrate.
Variousmixing configurationswere testedand itwas foundadditionof sodiumborohydrideat a fast
ratenearthestirbartipproducedthesmallestNPs(6.7±1.7nm).Ingeneral,largermorepolydispersed
NPs were found when the mixing conditions encouraged areas rich in silver nitrate and deficient in
sodiumborohydride.GoldNPswere synthesized using the Turkevichmethod,where tetrachloroauric
acidisreducedbytrisodiumcitrateatanelevatedtemperature.Usingthesamemixingconfigurationsas
those for the silverNP synthesis, the smallestNPs (13.1±2.2nm)wereobtainedwhen the reducing
Chapter6
131
agent(trisodiumcitrate)wasaddedatafastratenearthestirbartiptotheprecursor(tetrachloroauric
acid).Theseweresignificantlysmallerthanifthetrisodiumcitratewasaddedataslowratefromabove
thesolution(18±4.8nm),showingthatthemasstransferconditionsandhowthetrisodiumcitrateis
administeredtotheprecursorintheclassicalTurkevichmethodisimportant.Themixingconfiguration
didnotsignificantlyaffect thesizewhenusing the inversemethod, similar towhat isobserved in the
literature for similar citrate to gold molar ratios. However, fast addition of reagent produced lower
dispersitythanslowadditionbecauseitproducesasharperseparationofnucleationandgrowth.
The study highlights the importance of mass transfer conditions in determining the size and
dispersityoftheresultantNPs,evenwhenthemolarratioofreducingagenttoprecursorisfixed.Often,
batchreactorstudiesstatethemolarratioofsyntheses,butthismaybeinsufficienttoreproduceresults
sincespecificinformationonthemannerinwhichreagentsaremixedmayalsobeneeded.Theresults
demonstrate that both diffusion and advection mass transport mechanisms are important in
determiningtheresultantsizeanddispersityofsynthesisedNPs.Masstransportconceptsofadvection
anddiffusioncanexplainapparentdifferences insynthesesusingsimilarconcentrationof reagentbut
differentapparatus.
132
7 Conclusionsandfurtherwork
7.1 Conclusions
Theworkpresentedinthisthesisfocusesontheroleofmasstransferonthesizeanddispersityofsilver
and goldNPs synthesized inmicrofluidic devices. To achieve this, characterization of themicrofluidic
devices in termsofmass transfer characteristicswasmade tounderstandhowconcentrationprofiles
would evolve during a reaction. The concentrations at which reaction, nucleation and growth of NP
formationtakesplaceareveryimportantindeterminingtheirsizeanddispersity.Theobjectiveistobe
abletocontrolsizeanddispersityofsynthesizedNPssincethiscontrolstheirproperties,andhencetheir
suitabilityforuseinanyapplication.TheconcentrationatwhichtheprocessesforNPformationoccuris
controlledbymasstransfer.Subsequently,avarietyofmicrofluidicdevicesweretestedwithgoldand
silverNPsystemstoinvestigatehowchangingmasstransferconditionsaffectedthesizeanddispersity.
ThegoldNPsysteminvestigatedwasthewell-knownTurkevichmethod,wheretetrachloroauricacidis
reducedviacitrateatelevatedtemperatures.ThesilverNPsystemwasthereductionofsilvernitratevia
sodiumborohydrideinthepresenceofaligand(trisodiumcitrateorPVA).Themicrofluidicdevicesused
werethecoaxial flowreactor (CFR), impinging jet reactor (IJR),coiled flow inverter (CFI)andsplitand
recombine (SAR) mixer. A batch reactor study for the synthesis of silver and gold NPs was also
performed.
The primary objective of this study was to investigate how a variety of different mass transfer
conditionscanaffectthesizeanddispersityofsynthesisedNPs.Theoverarchingconclusionofthestudy,
afterinvestigatingavarietyofdifferentmasstransferconditionsusingmicrofluidicdevicesandabatch
reactorvessel,isthatmasstransferisanimportantpropertytoconsiderwhenmanipulationofsizeand
dispersityofNPsisdesired.ThevarietyofreactorsusedforthesynthesisofNPsinthisstudyeachhad
their own unique mass transport characteristics which can be summarised as follows: the CFR
maintained a separation of reagents through the channel because of the laminar flow and lack of
geometry manipulation to reduce diffusion distance, the IJR presented an efficient mixing device as
demonstratedbythelowmixingtimesfoundwiththeconditionstested(<10ms),theCFIutilisedDean
flowtoaidthemixingofthereagentswhileimprovingRTD,theSARalsopresentedanefficientmixing
device through channel geometry manipulation which reduces the diffusion distance and the batch
reactor vessel used external energy through the action of amagnetic stirrerwhich enhancedmixing
throughanadvectionmechanismtospeedupthebulkmixingofthefluid.
Tomanipulate the size of synthesised NPs, controlling reaction and subsequent nucleation and
growthisthemainconsideration.Theseprocessesaredrivenbytheconcentrationsatwhichtheytake
place,which is ultimately determined by themass transport. Another important consideration is the
stabilisationwhenachievingsmallmonodispersedNPs isdesired.This increases thecomplexityof the
problemofNPsizecontrol,particularlywhenthestabilisingagentbehavesasareducingagent(asisthe
casewiththereactionsstudiedinthiswork).TheCFRpresentsapromisingmicrofluidicdevicetocontrol
sizeofsynthesisedNPsbecauseofrestrictionofreactionataninterfacebetweenreagentstreams(due
Chapter7
133
tolaminarflowandsimplechannelgeometry).Thisallowsforareservoirofreagentswhicharesupplied
to the reaction interface at a constant concentration at a fixed point along the channel length. The
concentrationofreagentsdropsasreactionproceedsalongthelengthofthechannel,whichresultsina
naturaltransitionfromnucleationtogrowth.TheCFRalsosuppressesfoulingbecausethesynthesisof
nanoparticlesisprimarilytakingplaceawayfromchannelwalls.ThemajordrawbackoftheCFRisthatit
would require very long channels to allow reagents to fully mix (because mass transport through
diffusiononlyisaslowprocess),orthroughtheutilisationofsmallerchanneldimensions.Althoughitis
possibletoimplementtheCFRwithlongchannels,thismaybeunfeasiblebecauseofspacelimitations
within a laboratory, and the use of smaller channel dimensions would present the need for more
powerful (andexpensive)pumps.The IJRperformedextremelywellwhenconsidering theproblemof
foulingofchannelwallswhensynthesisingNPs.SincetherewerenochannelwallsintheIJR,fouling(or
cleaning)waseliminated. ItdemonstratedagoodcontroloverNPsizeand iseasilyscaledupthrough
numbering up. However, the operation window in terms of flow rates lacks flexibility because a
minimumflowrateisrequiredtoobtainjetswhiletoohighaflowratewouldresultinadeteriorationof
mixingquality.Ontopofthis,itisunknownhowtheflowrateratio(iftheflowratesfromeachjetare
notequal)wouldaffect themass transportproperties.TheCFIpresentsan improvedRTDbut fouling
wasobservedundercertainconditions,andtheSARsufferedthemostfromfoulingproblem(becauseof
ahighsurfaceareatovolumeratio)eventhoughitisanefficientmixer.Thismakesthesereactorsless
suitable for synthesis of NPs since they produce a dynamic condition as fouling evolves within the
channel over the course of the reaction. The batch reactor is a simple and effective means of
synthesisingNPs,whichisperhapsthebiggestadvantageinthesensethatitistheeasiesttoimplement
within the laboratory setting. It offers a quick and easyway to test a variety of synthesis conditions,
whichcanthenbe further investigatedwitha focusonmass transportwithin themicrofluidicdevices
presentedinthisworkaswellasotherspresentedintheliterature.Thelowsurfaceareatovolumeratio
withinthebatchreactorvesselsusedinthisworkminimisestheissueoffoulingbutneverthelessthere
wassomefoulingobservedonthePTFEmagneticstirrersusedinthesyntheses.Ofthedevicesstudied
inthiswork,theCFRandIJRpresentthemostpromisingreactortechnologiesforcontrollingthesizeof
NPsusingmicrofluidicdevices. It is hoped that this studyprovides aplatform for furtherwork in the
areaofsynthesisofnanomaterialsusingmicrofluidicdevices.
7.2 Furtherwork
Therearesomeexcitingdirections inwhichworkonsynthesisofNPs inmicrofluidicdevicescantake.
Ultimately,thegoalwouldbetoproducelargequantitiesofNPsinareliablemannerwiththeabilityto
tunethesizeconveniently.
TheCFRhasshownthepotentialtocontrolsizethroughvariationinresidencetimeusinglaminar
flow.AvortexflowregimeathigherReynoldsnumberwasalsotested,butdidn’tproducegoodresults
in termsof increasingcontroloversize.Furtherworkusing theCFRshould includeoperating itunder
turbulentconditionstoachievemixingkineticsfasterthanthereaction.Thiswouldintheoryallowfor
reproducible results sincemass transferwouldno longerplay a factorbecause concentrationswithin
thereactorwouldbehomogenizedbeforereactionoccurred. Astudyonthekineticsof thesilverNP
Conclusionsandfurtherwork
134
system to discover the reaction rate equation and kinetic rate constant would be beneficial to
determinehowefficientthemixingneedstobe.Theoreticallythisshouldincreasemonodispersitysince
reaction,nucleation,growthandstabilizationcantakeplaceatthesameconcentrationacrossthebulk
ofthefluid,ratherthanatvariouslocalizedconcentrationswithinthereactor.Discoveryofthereaction
rateequationandrateconstantwouldalsoenablethesimulationofaccurateCFDmodels,enablingthe
potentialtotestavarietyofreactordesignsinavirtualenvironmentratherthanhavingtofabricateand
test them experimentally. However, CFD can be computationally expensive for accuratemodels and
would be dependent on designing models which reduce computational expense or the use of very
powerfulcomputers.
TheIJRoffersexceptionalcontinuousflowcharacteristicsforsynthesisofNPs,namelythelackofa
possibilityofblockagethroughfouling.Becauseofthisthereisnoneedtocleanthereactorbeforeruns
withaquaregiaforinstance,meaningthatitcanproducevarioussamplesofNPsatdifferentflowrates
quickly. This makes it ideal to rapidly obtain information on new NP systems when the idea is to
investigatehowmasstransferaffectssizeanddispersity.OnelimitationoftheIJRhoweveristhatithas
anarrowrangeofoperability intermsofflowrates,becauseitrequiresacertainflowratetoactually
generatea jetand ithasa limiton themaximumflowrate since toohigha flow ratewill generatea
spraythatdoesnotmixefficiently.OneinterestingpossibilitywouldbetooperatetheIJRwithunequal
volumetricflowrateratios,to investigatehowthemasstransfer isaffected.HydrodynamicsoftheIJR
wouldbeaffected,withthecombinedstreambeingskewedinonedirectiondependentonwhichside
the jet velocity was higher. By knowing the mass transfer characteristics under these conditions,
operationunderunequalvolumetricflowratesforeachinputwouldallowanincreaseintheconditions
thatonecouldtestquicklyforanewNPsystem.
Flow reactors in general also afford benefits over batch reactors that would make them ideal
candidatesforinvestigatingkineticsandmechanismsofvariousNPsystems.Thisisbecauseofconcept
of space time ina flow reactorasopposed to reaction time inabatch reactor.More specifically, the
lengthor residence time ina reactorallows the reactions tobestudiedatanyparticularmoment.By
conductingin-linecharacterizationsuchasUV-VisorSAXS,onecanstudytheprocessofNPformationat
very small time resolutions (similar to the studiesbyPolteetal. andTakesueetal.)82, 121 This system
wouldalsoallowforawiderangeofconditionstobestudiedquicklywhenoperatedinconjunctionwith
a microfluidic device which can be operated continuously for long periods. This would help obtain
informationtounderstandthemechanismsofNPformationinarelativelyrapidmannerasopposedto
usingbatchreactors.
Synthesis of NPs in the batch reactor also revealed that different results can be obtained by
changing themass transferconditions.There isaneed formorespecific informationwhendescribing
how reagents aremixedwhen consideringbatch reactor synthesis studies.Batch reactors still offer a
verysimpleandeffectiveroutetosynthesizeNPs,andareagoodtooltoinformresearcherswhowant
to translate syntheses to flow. Itwould be interesting to see how the batch reactor results could be
reproduced in a flow system, for instance using a distributed feed to mimic the droplet addition of
reagenttothebulkinthebatchreactor.
Chapter7
135
136
AppendixA:MixingcharacterizationusingtheVillermaux-Dushmanreactions
Triiodideextinctioncoefficientevaluation
Aknownconcentrationof iodineismadebymixingastoichiometricbalanceof iodideandiodate ions
usingthefollowingconcentrations:
[KI]=1.13mM
[KIO3]=0.226mM
Tothisweaddanexcessofsulphuricacidtoconvertalltheiodideandiodateintoiodineaccordingto
thereaction:
𝐼𝑂QK + 5𝐼K + 6𝐻T → 3𝐼9 + 3𝐻9𝑂
14.2mMH2SO4 isaddedtoamixtureof2.26mMKIand0.452mMKIO3withequalvolumes(i.e. the
volumeofacidisequaltothatoftheiodide/iodatemixture).Thisresultsinaconcentrationaftermixing
of 7.1mMH2SO4, 1.13mMKI and 0.226mMof KIO3. By combining these reagents in the quantities
above,oneobtainsthefollowingiodineconcentration:
5 𝐼𝑂QK = 𝐼K =53𝐼9 = 1.13𝑚𝑀
𝐼9 = 0.68𝑚𝑀
Diluting the resulting solution to 20%, 40%, 60%, 80% and 100% the original concentration and
measuringtheabsorbanceat460nm(wavelengthatwhich iodineabsorbs lightsignificantly),onecan
obtain theextinction coefficient for iodine. FigureA1 shows theobtained calibration curve for iodine
fromwhich an extinction coefficient of 626 l/mol.cm is obtained (literature value of 758 l/mol.cm is
reportedforiodineat460nm).189
137
FigureA1:Calibrationcurvefor iodine,absorbancevs.concentration,withstraight lineusedtoobtain
theextinctioncoefficientofiodine.
Followingthis,wepreparesolutionswiththenecessaryconcentrationfortriiodideextinctioncoefficient
calculations. To do this the saturationmethod is used. In thismethod, we take a known amount of
iodineandmixwithanexcessofiodideasperthefollowingreaction:
𝐼K + 𝐼9 ↔ 𝐼QK
By using a large excess of iodide ions (saturation), the equilibrium is driven largely in favor of the
triiodideion.Thus,1mlof0.321MKIwasaddedto1mlof92.88µMiodine.Thepreparedsolutionis
thendilutedtotheappropriateconcentrationsandtheabsorbanceismeasuredat353nmtoobtainthe
calibration curve and extinction coefficient for triiodide. A value of 23209 l/mol.cm is found for the
triiodideextinctioncoefficientshownFigureA2.
y=626.07x
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0.00E+00 2.00E-04 4.00E-04 6.00E-04 8.00E-04
Absorban
ce@
460nm
Iodineconcentration(M)
138
FigureA2:Calibrationcurvefortriiodide,absorbancevs.concentration,withstraightlineusedtoobtain
theextinctioncoefficientoftriiodide.
y=23209x
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0.00E+00 2.00E-05 4.00E-05 6.00E-05
Absorban
ce@
353nm
Triiodideconcentration(M)
139
gProms model for solution of differential mass balance equations for the Villermaux-Dushman
reactions:
Model:
PARAMETER
AlphaASREAL#Volumetricfractionofacid
k1ASREAL#reactionrateconstantforreaction1,l/mol.s
k3ASREAL#forwardreactionrateconstantforreaction3,l/mol.s
k3primeASREAL#backwardreactionrateconstantforreaction3,1/s
kASINTEGERDEFAULT6#chemicalspecies(1=H+,2=H2BO3-,3=I-,4=IO3-,5=I2,6=I3-)
sASINTEGERDEFAULT2#streamnumber(1=acidstream,2=reagentstream)
tmASREAL#mixingtime,s
VARIABLE
k2ASNoType#Reactionrateconstantforreaction2,whichvarieswithionic
strength,l/mol.s
CASARRAY(k,s)OFMolarConcentration#Concentrationofcomponentkinstreams,mol/l
CaASARRAY(k)OFNoType#Averageconcentrationofcomponentk,mol/l
r1ASARRAY(s)OFReactionRate#Reactionrateforreaction1instreams,mol/l.s
r2ASARRAY(s)OFReactionRate#Reactionrateforreaction2instreams,mol/l.s
r3ASARRAY(s)OFReactionRate#Reactionrateforreaction3instreams,mol/l.s
IASMolarConcentration#ionicstrength
EQUATION
#DifferentialMassbalanceequationsforeachspecies
#chemicalspecies(1=H+,2=H2BO3-,3=I-,4=IO3-,5=I2,6=I3-)
FORm:=1TOsDO
$C(1,m)=(Ca(1)-C(1,m))/tm-r1(m)-6*r2(m);
$C(2,m)=(Ca(2)-C(2,m))/tm-r1(m);
$C(3,m)=(Ca(3)-C(3,m))/tm-5*r2(m)-r3(m);
$C(4,m)=(Ca(4)-C(4,m))/tm-r2(m);
$C(5,m)=(Ca(5)-C(5,m))/tm+3*r2(m)-r3(m);
$C(6,m)=(Ca(6)-C(6,m))/tm+r3(m);
END
#averageconcentrationofspecies
FORj:=1TOkDO
Ca(j)=(1-alpha)*C(j,2)+alpha*C(j,1);
END
140
#reactionrateequations
FORz:=1TOsDO
r1(z)=k1*C(1,z)*C(2,z);
r2(z)=k2*C(1,z)^2*C(3,z)^2*C(4,z);
r3(z)=k3*C(5,z)*C(3,z)-k3prime*C(6,z);
END
Process:
UNIT
IEMASIEM_VMD
SET
WITHINIEMDO
tm:=1e-4;#mixingtime,s
k1:=1e20;#reactionrateconstantforreaction1,l/mol.s
k3:=5.6e9;#forwardreactionrateconstantforreaction3,l/mol.s
k3prime:=7.5e6;#backwardreactionrateconstantforreaction3,1/s
Alpha:=0.5;#Volumetricfractionofacid
END
EQUATION
#Equations
WITHINIEMDO
I=0.5*(C(1,1)+C(1,2)+C(2,1)+C(2,2)+C(3,1)+C(3,2)+C(4,1)+C(4,2)+C(6,1)+C(6,2));
k2 = 10^(9.28-3.66*I^0.5); #Reaction rate constant for reaction 2, which varies with ionic strength,
l/mol.s
END
INITIAL
WITHINIEMDO
C(1,1)=0.04; #acidconcentration
FORa:=2TOkDO #setsallotherspeciesconcentrationtozero
C(a,1)=0;
END
#chemicalspecies(1=H+,2=H2BO3-,3=I-,4=IO3-,5=I2,6=I3-),
#setconcentrationforbuffer
#conc.mol/l
C(1,2)=0;
C(2,2)=0.09/2;
C(3,2)=0.032;
141
C(4,2)=0.006;
C(5,2)=0;
C(6,2)=0;
END
SOLUTIONPARAMETERS
ReportingInterval:=1e-3;
SCHEDULE
#OperationSchedule
Continuefor10;
142
AppendixB:Synthesisofsilvernanoparticlesinamicrofluidiccoaxialflowreactor
FigureB1:Microscopeimageshowingthefoulingofthechannelinasplitandrecombinemicromixeraftersynthesis
ofsilvernanoparticlesusingsilvernitrate,sodiumborohydrideandtrisodiumcitrate.
FigureB2:PeakabsorbanceofsilverNPssynthesizedatvarioustotalflowratesintherange1-14ml/min.Synthesis
was repeated four times for each flow rate. Concentrations of silver nitrate 0.2mM, trisodium citrate 0.2mM,
sodiumborohydride0.3mM.0.798mminnertubeI.D.
143
Figure B3: UV-vis spectra of silver NPs synthesized at various total flow rates in the range 1-14 ml/min.
Concentrationsof silvernitrate0.2mM, trisodiumcitrate0.2mM,sodiumborohydride0.3mM.0.798mm inner
tubeI.D.
FigureB4:Foulingontheinnerwalloftheinnertubewhensilvernitratesolutionflowedthroughtheinnertube
Fig.B5:Foulingontheouterwalloftheinnertubewhensilvernitratesolutionflowedthroughtheoutertube
144
AppendixC:Synthesisofsilvernanoparticlesusingamicrofluidicimpingingjet
reactor
Reproducibilityofsynthesis
The reproducibilityof the syntheses in the IJRwas investigatedby repeatingexperimentswithcitrate
and PVA. The peak absorbance obtained from UV-Vis spectroscopy was used as a measure for the
reproducibility. Figure C1 shows reproducibility for the synthesis in the 0.5mm I.D. tubing IJR using
citrate as a ligand. FigureC2 shows reproducibility for the synthesis in the0.25mmand0.5mm I.D.
tubingIJRusingPVAasaligand.
Figure C1: Peak absorbance (peak wavelength varied between 386-389 nm) obtained from UV-Vis
spectroscopyvs.WebernumberofsilverNPssynthesizedatdifferenttotalflowratesusing0.5mmI.D.
tubingIJR.Experimentsateachflowratewererepeatedthreetimes.Concentrationofsilvernitrate0.9
mM,trisodiumcitrate6mMininput1andsodiumborohydride1.8mMininput2.
145
Figure C2: Peak absorbance (peak wavelength varied between 398-401 nm) obtained from UV-Vis
spectroscopyvs.WebernumberofsilverNPssynthesizedatdifferenttotalflowratesusing: i)0.5mm
(black squares) and ii) 0.25 mm (red triangles) I.D. tubing IJR. Each flow rate was repeated twice.
Concentrationof silver nitrate 0.9mM, PVA0.02wt% in input 1 and sodiumborohydride 1.8mM in
input2.
146
ComparisonofhydrodynamicsbetweenwateronlysystemandsilverNPsynthesisusingPVAin0.25
mmIJR
FigureC3comparesthehydrodynamicsoftheimpingingjetsinthe0.25mmI.D.tubingIJRwhenwater
isusedandforthesilverNPsynthesiswhereoneofthestreamscontainsPVA.
FigureC3:Highspeedcameraimagesofhydrodynamicsinthe0.25mmI.D.tubingIJRforasystemusingonlywater
and under silver NP synthesis conditions of concentration of silver nitrate 0.9 mM, PVA 0.02 wt%, sodium
borohydride1.8mM.Totalflowrateswere:A:18ml/min(We:32),B:22ml/min(We:48),C:26ml/min(We:68),
D:30ml/min(We:90),E:34ml/min(We:115),F:38ml/min(We:144)andG:42ml/min(We:176).
147
TEMimageofsilverNPssynthesizedusinganincreasedPVAconcentration
FigureC4showsthesilverNPsobtainedwhenthePVAconcentrationwasdoubled.
FigureC4:TEMimageandparticlesizedistributionofsilverNPssynthesizedusingthe0.5mmI.D.tubingIJRata
flow rateof72ml/min (We:65).Concentrationof silvernitrate0.9mM,PVA0.04wt%, sodiumborohydride1.8
mM.
148
AppendixD:Effectofhydrodynamicsandmixingconditionsonthecontinuous
synthesisofgoldandsilvernanoparticlesinacoaxialflowdevice
GoldNPsynthesisrepeatability
FigureD1:Averagediameterof goldNPs synthesizedusing theCFRwithaCFI residence loopat flow
ratesof0.5,1and3ml/minobtainedusingDCSanalysis.Concentrationoftetrachloroauricacid,0.557
mM; concentration of trisodium citrate, 0.09M; volumetric flow rate ratio, 32.3: 1 (Qout:Qin, HAuCl4:
Na3citrate);molarflowrateratio,1:5(HAuCl4:Na3citrate);temperature,80°C.
149
Figure D2: Average diameter of gold NPs synthesized using the CFR with a CFI residence loop at
temperatures between 60-100°C obtained using DCS analysis. Concentration of tetrachloroauric acid,
0.557 mM; concentration of trisodium citrate, 0.09 M; volumetric flow rate ratio, 32.3: 1 (Qout:Qin,
HAuCl4:Na3citrate);molarflowrateratio,1:5(HAuCl4:Na3citrate);flowrate,1ml/min.
150
VisualizationofflowintheCFRusingturbulentconditions
Figure D3: Flow visualization in the CFR with an inner tube I.D. of 0.798 mm at different Reynolds
numbersa)5.1ml/min(Reinnertube,133,Remainchannel,54),b)10.1ml/min(Reinnertube,265,
Remainchannel,107)andc)15.1ml/min (Re inner tube,397,Remainchannel,160).BasicBluedye
was pumped through the inner tube between 5 and 15ml/min andwaterwas pumped through the
outertubeatafixedflowrateof0.1ml/min.1indicateswhendyeisinitiallypumpedthroughtheinner
tubewithincreasingnumbersgoinguptoshowtheCFRatsteadystate.
151
SilverNPsynthesisrepeatabilitywithinnertubeI.D.variation
FigureD4: Peak absorbance obtained fromUV-Vis spectra vs. inner tube internal diameter of repeat
experiments for silver NPs synthesized using the CFR with inner tube internal diameters ranging
between 0.142 and 0.798 mm. Concentration of silver nitrate, 0.1 mM; concentration of sodium
borohydride,0.3mM;concentrationoftrisodiumcitrate,0.5M;volumetricflowrateratio,1:1(Qout:Qin,
NaBH4:AgNO3);molarflowrateratio,1:3:5(AgNO3:NaBH4:Na3citrate);flowrate,1ml/min,temperature
22-24°C.
152
Listofpublications
Thefollowinglistarepublicationswhichresultedfromtheworkcontainedinthisthesis:
1. R.Baber,L.Mazzei,N.T.K.ThanhandA.Gavriilidis,RSCAdvances,2015,5,95585-95591.
2. R.Baber,L.Mazzei,N.T.ThanhandA.Gavriilidis,JournalofFlowChemistry,2016,1-11.
153
Bibliography
1. A.Gavriilidis,P.Angeli,E.Cao,K.K.YeongandY.S.S.Wan,ChemicalEngineeringResearch&
Design,2002,80,3-30.
2. V.V.Mody,R.Siwale,A.SinghandH.R.Mody,JournalofPharmacyandBioalliedSciences,
2010,2,282.
3. W.K.Leutwyler,S.L.BürgiandH.Burgl,Science,1996,271,933.
4. W.C.Chan,D.J.Maxwell,X.Gao,R.E.Bailey,M.HanandS.Nie,Currentopinionin
biotechnology,2002,13,40-46.
5. S.Agnihotri,S.MukherjiandS.Mukherji,RSCAdvances,2014,4,3974-3983.
6. Y.SunandY.Xia,AdvancedMaterials,2003,15,695-699.
7. M.Maillard,S.GiorgioandM.-P.Pileni,AdvancedMaterials,2002,14,1084.
8. B.D.Busbee,S.O.ObareandC.J.Murphy,AdvancedMaterials,2003,15,414-416.
9. P.Hugounenq,M.Levy,D.Alloyeau,L.Lartigue,E.Dubois,V.Cabuil,C.Ricolleau,S.Roux,C.
Wilhelm,F.GazeauandR.Bazzi,TheJournalofPhysicalChemistryC,2012,116,15702-15712.
10. O.M.Bakr,B.H.WunschandF.Stellacci,ChemistryofMaterials,2006,18,3297-3301.
11. H.Khurshid,S.Chandra,W.Li,M.H.Phan,G.C.Hadjipanayis,P.MukherjeeandH.Srikanth,J
ApplPhys,2013,113,17B508.
12. C.J.MurphyandN.R.Jana,AdvancedMaterials,2002,14,80.
13. E.V.Shevchenko,M.I.Bodnarchuk,M.V.Kovalenko,D.V.Talapin,R.K.Smith,S.Aloni,W.
HeissandA.P.Alivisatos,Adv.Mater,2008,20,4323-4329.
14. A.D.McFarlandandR.P.VanDuyne,Nanoletters,2003,3,1057-1062.
15. G.Korotcenkov,B.K.Cho,L.B.GulinaandV.P.Tolstoy,SensorsandActuatorsB:Chemical,
2012,166–167,402-410.
16. N.Bahadur,K.Jain,R.Pasricha,GovindandS.Chand,SensorsandActuatorsB:Chemical,2011,
159,112-120.
17. D.D.EvanoffandG.Chumanov,ChemPhysChem,2005,6,1221-1231.
18. S.Panigrahi,S.Praharaj,S.Basu,S.K.Ghosh,S.Jana,S.Pande,T.Vo-Dinh,H.JiangandT.Pal,
TheJournalofPhysicalChemistryB,2006,110,13436-13444.
19. C.J.Murphy,T.K.Sau,A.M.Gole,C.J.Orendorff,J.Gao,L.Gou,S.E.HunyadiandT.Li,The
JournalofPhysicalChemistryB,2005,109,13857-13870.
20. D.Chen,X.Qiao,X.QiuandJ.Chen,JMaterSci,2009,44,1076-1081.
21. Y.Li,Y.WuandB.S.Ong,JAmChemSoc,2005,127,3266-3267.
22. Z.-J.Jiang,C.-Y.LiuandL.-W.Sun,TheJournalofPhysicalChemistryB,2005,109,1730-1735.
23. A.C.Patel,S.Li,C.Wang,W.ZhangandY.Wei,Chemistryofmaterials,2007,19,1231-1238.
24. S.Link,Z.L.WangandM.El-Sayed,TheJournalofPhysicalChemistryB,1999,103,3529-3533.
25. C.Marambio-JonesandE.M.Hoek,JNanopartRes,2010,12,1531-1551.
26. G.A.SotiriouandS.E.Pratsinis,CurrentOpinioninChemicalEngineering,2011,1,3-10.
154
27. V.K.Sharma,R.A.YngardandY.Lin,AdvancesinColloidandInterfaceScience,2009,145,83-
96.
28. Y.Yang,S.Matsubara,L.Xiong,T.HayakawaandM.Nogami,TheJournalofPhysicalChemistry
C,2007,111,9095-9104.
29. R.A.Sperling,P.RiveraGil,F.Zhang,M.ZanellaandW.J.Parak,Chem.Soc.Rev.,2008,37,
1896-1908.
30. L.DykmanandN.Khlebtsov,Chem.Soc.Rev.,2012,41,2256-2282.
31. S.GuoandE.Wang,AnalyticaChimicaActa,2007,598,181-192.
32. M.-C.DanielandD.Astruc,ChemicalReviews,2004,104,293-346.
33. W.Ehrfeld,V.HesselandH.Lowe,Microreactors:NewTechnologyforModernChemistry,
Wiley-VCH,Weinheim,2000.
34. N.T.K.Thanh,N.MacleanandS.Mahiddine,ChemicalReviews,2014,114,7610–7630.
35. A.Abou-Hassan,O.Sandre,S.NeveuandV.Cabuil,AngewandteChemieInternationalEdition,
2009,48,2342-2345.
36. A.Abou-Hassan,R.BazziandV.Cabuil,AngewandteChemieInternationalEdition,2009,48,
7180-7183.
37. S.A.Khan,A.Günther,M.A.SchmidtandK.F.Jensen,Langmuir,2004,20,8604-8611.
38. X.Shen,Y.Song,S.Li,R.Li,S.Ji,Q.Li,H.Duan,R.Xu,W.YangandK.Zhao,RSCAdvances,2014,
4,34179-34188.
39. R.Wang,W.Yang,Y.Song,X.Shen,J.Wang,X.Zhong,S.LiandY.Song,Scientificreports,2015,
5,9189.
40. W.Chen-Hsun,H.Chih-Chia,Y.Chen-Sheng,L.Huan-YaoandL.Gwo-Bin,JMicromech
Microeng,2008,18,035019.
41. J.M.Ottino,AnnuRevFluidMech,1990,22,207-253.
42. N.T.NguyenandZ.G.Wu,JMicromechMicroeng,2005,15,R1-R16.
43. C.J.CampbellandB.A.Grzybowski,PhilosophicalTransactionsoftheRoyalSocietya-
MathematicalPhysicalandEngineeringSciences,2004,362,1069-1086.
44. E.A.Mansur,M.Ye,Y.WangandY.Dai,ChineseJournalofChemicalEngineering,2008,16,
503-516.
45. D.Bokenkamp,A.Desai,X.Yang,Y.C.Tai,E.M.MarzluffandS.L.Mayo,AnalyticalChemistry,
1998,70,232-236.
46. N.Kockmann,T.Kiefer,M.EnglerandP.Woias,2005.
47. S.H.Wong,M.C.L.WardandC.W.Wharton,Sens.ActuatorB-Chem.,2004,100,359-379.
48. M.YiandH.H.Bau,InternationalJournalofHeatandFluidFlow,2003,24,645-656.
49. F.G.Bessoth,A.J.deMelloandA.Manz,AnalyticalCommunications,1999,36,213-215.
50. J.B.Knight,A.Vishwanath,J.P.BrodyandR.H.Austin,PhysRevLett,1998,80,3863-3866.
51. F.Schonfeld,V.HesselandC.Hofmann,LabonaChip,2004,4,65-69.
52. M.Kashid,A.RenkenandL.Kiwi-Minsker,ChemEngJ,2011,167,436-443.
53. W.Buchegger,C.Wagner,B.Lendl,M.KraftandM.J.Vellekoop,MicrofluidNanofluid,2011,
10,889-897.
155
54. Z.Chen,M.Bown,B.O’Sullivan,J.MacInnes,R.Allen,M.Mulder,M.BlomandR.van’tOever,
MicrofluidNanofluid,2009,6,763-774.
55. S.WaelchliandP.RudolfvonRohr,InternationalJournalofMultiphaseFlow,2006,32,791-
806.
56. M.N.Kashid,D.F.Rivas,D.W.AgarandS.Turek,Asia-PacificJournalofChemicalEngineering,
2008,3,151-160.
57. W.Tanthapanichakoon,N.Aoki,K.MatsuyamaandK.Mae,ChemEngSci,2006,61,4220-4232.
58. R.Miyake,T.S.J.Lammerink,M.ElwenspoekandJ.H.J.Fluitman,Micromixerwithfast
diffusion,1993.
59. R.Miyake,K.Tsuzuki,T.Takagi,K.ImaiandIeee,inMems97,Proceedings-IeeetheTenth
AnnualInternationalWorkshoponMicroElectroMechanicalSystems:AnInvestigationofMicro
Structures,Sensors,Actuators,MachinesandRobots,1997,DOI:
10.1109/memsys.1997.581779,pp.102-107.
60. J.Voldman,M.L.GrayandM.A.Schmidt,JournalofMicroelectromechanicalSystems,2000,9,
295-302.
61. V.Hessel,H.LoweandF.Schonfeld,ChemEngSci,2005,60,2479-2501.
62. A.D.Stroock,S.K.W.Dertinger,A.Ajdari,I.Mezic,H.A.StoneandG.M.Whitesides,Science,
2002,295,647-651.
63. T.G.KangandT.H.Kwon,JMicromechMicroeng,2004,14,891-899.
64. J.A.Creighton,C.G.BlatchfordandM.G.Albrecht,JournaloftheChemicalSociety-Faraday
Transactionsii,1979,75,790-798.
65. N.Shirtcliffe,U.NickelandS.Schneider,J.ColloidInterfaceSci.,1999,211,122-129.
66. B.H.Ryu,Y.Choi,H.S.Park,J.H.Byun,K.Kong,J.O.LeeandH.Chang,ColloidSurfaceA,2005,
270,345-351.
67. K.Song,S.Lee,T.ParkandB.Lee,KoreanJ.Chem.Eng.,2009,26,153-155.
68. J.Polte,X.Tuaev,M.Wuithschick,A.Fischer,A.F.Thuenemann,K.Rademann,R.Kraehnert
andF.Emmerling,ACSNano,2012,6,5791-5802.
69. V.V.Pinto,M.J.Ferreira,R.Silva,H.A.Santos,F.SilvaandC.M.Pereira,ColloidsandSurfaces
A:PhysicochemicalandEngineeringAspects,2010,364,19-25.
70. A.Thogersen,J.Bonsak,C.H.FosliandG.Muntingh,JApplPhys,2011,110.
71. J.Wagner,T.R.TshikhudoandJ.M.Koehler,ChemEngJ,2008,135,S104-S109.
72. J.M.Kohler,L.Abahmane,J.Wagner,J.AlbertandG.Mayer,ChemEngSci,2008,63,5048-
5055.
73. X.Z.Lin,A.D.TerepkaandH.Yang,NanoLetters,2004,4,2227-2232.
74. C.Sonnichsen,J.Boleininger,A.KurzandV.Reuss,PhysChemChemPhys,2006,8,3824-3827.
75. N.R.Jana,L.GearheartandC.J.Murphy,Chem.Commun.,2001,DOI:10.1039/b100521i,617-
618.
76. A.Knauer,A.Thete,S.Li,H.Romanus,A.Csaki,W.FritzscheandJ.M.Koehler,ChemEngJ,
2011,166,1164-1169.
77. J.D.Tice,H.Song,A.D.LyonandR.F.Ismagilov,Langmuir,2003,19,9127-9133.
156
78. J.M.Koehler,M.Held,U.HuebnerandJ.Wagner,ChemEngTechnol,2007,30,347-354.
79. H.Liu,J.Huang,D.Sun,L.Lin,W.Lin,J.Li,X.Jiang,W.WuandQ.Li,ChemEngJ,2012,209,
568-576.
80. D.V.R.Kumar,B.L.V.PrasadandA.A.Kulkarni,ChemEngJ,2012,192,357-368.
81. L.L.Lazarus,C.T.Riche,B.C.Marin,M.Gupta,N.MalmstadtandR.L.Brutchey,ACSApplied
Materials&Interfaces,2012,4,3077-3083.
82. M.Takesue,T.Tomura,M.Yamada,K.Hata,S.KuwamotoandT.Yonezawa,JAmChemSoc,
2011,133,14164-14167.
83. A.HengleinandM.Giersig,ChemInform,2000,31,9533-9539.
84. Z.S.PillaiandP.V.Kamat,TheJournalofPhysicalChemistryB,2003,108,945-951.
85. W.Zhang,X.Qiao,Q.Chen,Y.CaiandH.Chen,AppliedSurfaceScience,2012,258,5909-5913.
86. T.J.Woehl,J.E.Evans,I.Arslan,W.D.RistenpartandN.D.Browning,ACSNano,2012,6,8599-
8610.
87. X.C.Jiang,C.Y.Chen,W.M.ChenandA.B.Yu,Langmuir,2010,26,4400-4408.
88. J.Turkevich,P.C.StevensonandJ.Hillier,DiscussionsoftheFaradaySociety,1951,11,55-75.
89. G.Frens,Nature,1973,241,20-22.
90. S.L.Goodman,G.M.Hodges,L.K.TrejdosiewiczandD.C.Livingston,JournalofMicroscopy,
1981,123,201-213.
91. M.K.ChowandC.F.Zukoski,J.ColloidInterfaceSci.,1994,165,97-109.
92. C.-H.Su,P.-L.WuandC.-S.Yeh,TheJournalofPhysicalChemistryB,2003,107,14240-14243.
93. L.Pei,K.MoriandM.Adachi,Langmuir,2004,20,7837-7843.
94. J.Kimling,M.Maier,B.Okenve,V.Kotaidis,H.BallotandA.Plech,TheJournalofPhysical
ChemistryB,2006,110,15700-15707.
95. X.Ji,X.Song,J.Li,Y.Bai,W.YangandX.Peng,JAmChemSoc,2007,129,13939-13948.
96. W.PatungwasaandJ.H.Hodak,MaterialsChemistryandPhysics,2008,108,45-54.
97. I.Ojea-Jiménez,F.M.Romero,N.G.BastúsandV.Puntes,TheJournalofPhysicalChemistryC,
2010,114,1800-1804.
98. M.A.Uppal,A.Kafizas,T.H.LimandI.P.Parkin,NewJournalofChemistry,2010,34,1401-
1407.
99. M.A.Uppal,A.Kafizas,M.B.EwingandI.P.Parkin,NewJournalofChemistry,2010,34,2906-
2914.
100. I.Ojea-Jiménez,N.G.BastúsandV.Puntes,TheJournalofPhysicalChemistryC,2011,115,
15752-15757.
101. I.Ojea-JiménezandJ.M.Campanera,TheJournalofPhysicalChemistryC,2012,116,23682-
23691.
102. S.K.Sivaraman,S.KumarandV.Santhanam,J.ColloidInterfaceSci.,2011,361,543-547.
103. F.Schulz,T.Homolka,N.G.Bastús,V.Puntes,H.WellerandT.Vossmeyer,Langmuir,2014,30,
10779-10784.
104. M.Wuithschick,A.Birnbaum,S.Witte,M.Sztucki,U.Vainio,N.Pinna,K.Rademann,F.
Emmerling,R.KraehnertandJ.Polte,ACSNano,2015,9,7052-7071.
157
105. J.Piella,N.G.BastúsandV.Puntes,ChemistryofMaterials,2016,28,1066-1075.
106. F.Kettemann,A.Birnbaum,S.Witte,M.Wuithschick,N.Pinna,R.Kraehnert,K.Rademannand
J.Polte,ChemistryofMaterials,2016,28,4072-4081.
107. J.Ftouni,M.Penhoat,A.Addad,E.Payen,C.RolandoandJ.-S.Girardon,Nanoscale,2012,4,
4450-4454.
108. K.Sugano,Y.Uchida,O.Ichihashi,H.Yamada,T.TsuchiyaandO.Tabata,MicrofluidNanofluid,
2010,9,1165-1174.
109. J.M.Kohler,J.WagnerandJ.Albert,JournalofMaterialsChemistry,2005,15,1924-1930.
110. D.Shalom,R.C.R.Wootton,R.F.Winkle,B.F.Cottam,R.Vilar,A.J.deMelloandC.P.Wilde,
MaterialsLetters,2007,61,1146-1150.
111. V.SebastianCabeza,S.Kuhn,A.A.KulkarniandK.F.Jensen,Langmuir,2012,28,7007-7013.
112. H.Tsunoyama,N.IchikuniandT.Tsukuda,Langmuir,2008,24,11327-11330.
113. V.K.LaMerandR.H.Dinegar,JournaloftheAmericanChemicalSociety,1950,72,4847-4854.
114. J.Turkevich,P.C.StevensonandJ.Hillier,TheJournalofPhysicalChemistry,1953,57,670-673.
115. S.Biggs,M.K.Chow,C.F.ZukoskiandF.Grieser,J.ColloidInterfaceSci.,1993,160,511-513.
116. S.Biggs,P.Mulvaney,C.F.ZukoskiandF.Grieser,JournaloftheAmericanChemicalSociety,
1994,116,9150-9157.
117. J.F.Wall,F.Grieserandh.F.Zukoski,JournaloftheChemicalSociety,FaradayTransactions,
1997,93,4017-4020.
118. S.Kumar,K.S.GandhiandR.Kumar,Industrial&EngineeringChemistryResearch,2006,46,
3128-3136.
119. B.-K.Pong,H.I.Elim,J.-X.Chong,W.Ji,B.L.TroutandJ.-Y.Lee,TheJournalofPhysical
ChemistryC,2007,111,6281-6287.
120. J.r.Polte,T.T.Ahner,F.Delissen,S.Sokolov,F.Emmerling,A.ThünemannandR.Kraehnert,
JournaloftheAmericanChemicalSociety,2010,132,1296-1301.
121. J.Polte,R.Erler,A.F.Thunemann,S.Sokolov,T.T.Ahner,K.Rademann,F.EmmerlingandR.
Kraehnert,ACSNano,2010,4,1076-1082.
122. D.GoiaandE.Matijević,ColloidsandSurfacesA:PhysicochemicalandEngineeringAspects,
1999,146,139-152.
123. J.Park,V.PrivmanandE.Matijević,TheJournalofPhysicalChemistryB,2001,105,11630-
11635.
124. J.Polte,CrystEngComm,2015,17,6809-6830.
125. A.Plech,V.Kotaidis,A.SiemsandM.Sztucki,PhysChemChemPhys,2008,10,3888-3894.
126. A.Knauer,A.Csáki,F.Möller,C.Hühn,W.FritzscheandJ.M.Köhler,TheJournalofPhysical
ChemistryC,2012,116,9251-9258.
127. J.WagnerandJ.M.Kohler,NanoLetters,2005,5,685-691.
128. M.Takagi,T.Maki,M.MiyaharaandK.Mae,ChemEngJ,2004,101,269-276.
129. T.Maki,J.I.KitadaandK.Mae,ChemEngTechnol,2013,36,1027-1032.
130. A.AbouHassan,O.Sandre,V.CabuilandP.Tabeling,Chem.Commun.,2008,DOI:
10.1039/B719550H,1783-1785.
158
131. J.-M.Lim,A.Swami,L.M.Gilson,S.Chopra,S.Choi,J.Wu,R.Langer,R.KarnikandO.C.
Farokhzad,ACSNano,2014,8,6056-6065.
132. J.Schülein,I.Minrath,R.PommersheimandH.Löwe,JournalofFlowChemistry,2014,4,44-53.
133. J.Shen,Z.Li,Q.YanandY.Chen,TheJournalofPhysicalChemistry,1993,97,8504-8511.
134. W.M.Haynes,CRChandbookofchemistryandphysics,CRCPress;Taylor&Francis,BocaRaton,
Fla.[u.a.];London,95thedn.,2014.
135. J.C.Fanning,B.C.Brooks,A.B.Hoeglund,D.A.PelletierandJ.A.Wadford,InorganicaChimica
Acta,2000,310,115-119.
136. D.L.VanHyning,W.G.KlempererandC.F.Zukoski,Langmuir,2001,17,3120-3127.
137. V.R.Fernandes,A.M.F.R.PintoandC.M.Rangel,InternationalJournalofHydrogenEnergy,
2010,35,9862-9868.
138. H.I.Schlesinger,H.C.Brown,A.E.Finholt,J.R.Gilbreath,H.R.HoekstraandE.K.Hyde,JAm
ChemSoc,1953,75,215-219.
139. D.L.VanHyningandC.F.Zukoski,Langmuir,1998,14,7034-7046.
140. O.Brand,G.K.Fedder,C.Hierold,J.G.Korvink,O.TabataandN.Kockmann,Microprocess
engineering:fundamentals,devices,fabrication,andapplications,JohnWiley&Sons,
Weinheim,Germany,2013.
141. H.S.HarnedandC.L.Hildreth,JAmChemSoc,1951,73,3292-3293.
142. M.Chatenet,M.B.Molina-Concha,N.El-Kissi,G.ParrourandJ.P.Diard,ElectrochimicaActa,
2009,54,4426-4435.
143. M.Wuithschick,B.Paul,R.Bienert,A.Sarfraz,U.Vainio,M.Sztucki,R.Kraehnert,P.Strasser,K.
Rademann,F.EmmerlingandJ.Polte,ChemistryofMaterials,2013,25,4679-4689.
144. W.L.W.Hau,D.W.Trau,N.J.Sucher,M.WongandY.Zohar,JMicromechMicroeng,2003,13,
272-278.
145. R.Baber,L.Mazzei,N.T.K.ThanhandA.Gavriilidis,RSCAdvances,2015,5,95585-95591.
146. C.H.Sohn,W.S.Seol,A.A.ShibanovandV.P.Pikalov,JournalofPropulsionandPower,2007,
23,131-139.
147. A.J.MahajanandD.J.Kirwan,AIChEJournal,1996,42,1801-1814.
148. J.W.Mullin,Crystallization,Butterworth-Heinemann,Oxford,2001.
149. P.ErniandA.Elabbadi,Langmuir,2013,29,7812-7824.
150. D.R.Kumar,B.L.PrasadandA.A.Kulkarni,Ind.Eng.Chem.Res.,2013,52,17376-17382.
151. M.Hosni,S.Farhat,I.Hinkov,M.BenAmar,A.KanaevandN.Jouini,AIChEJournal,2015.
152. B.K.JohnsonandR.K.Prud'homme,AustralianJournalofChemistry,2003,56,1021-1024.
153. B.K.JohnsonandR.K.Prud’homme,PhysRevLett,2003,91,118302.
154. D.L.Marchisio,L.RivautellaandA.A.Barresi,AIChEJournal,2006,52,1877-1887.
155. H.C.SchwarzerandW.Peukert,ChemEngTechnol,2002,25,657-661.
156. S.W.Siddiqui,Y.Zhao,A.KukukovaandS.M.Kresta,Ind.Eng.Chem.Res.,2009,48,7945-
7958.
157. J.Aubin,M.FerrandoandV.Jiricny,ChemEngSci,2010,65,2065-2093.
158. V.Hessel,S.Hardt,H.LöweandF.Schönfeld,AIChEJournal,2003,49,566-577.
159
159. P.GuichardonandL.Falk,ChemEngSci,2000,55,4233-4243.
160. J.-M.CommengeandL.Falk,ChemicalEngineeringandProcessing:ProcessIntensification,
2011,50,979-990.
161. L.FalkandJ.M.Commenge,ChemEngSci,2010,65,405-411.
162. J.VillermauxandL.Falk,ChemEngSci,1994,49,5127-5140.
163. V.Bergeron,D.Bonn,J.Y.MartinandL.Vovelle,Nature,2000,405,772-775.
164. J.W.BushandA.E.Hasha,Journaloffluidmechanics,2004,511,285-310.
165. N.BremondandE.Villermaux,JournalofFluidMechanics,2006,549,273-306.
166. N.Ashgriz,W.BrocklehurstandD.Talley,JournalofPropulsionandPower,2001,17,736-749.
167. D.Paramelle,A.Sadovoy,S.Gorelik,P.Free,J.HobleyandD.G.Fernig,Analyst,2014,139,
4855-4861.
168. A.BhattacharyaandP.Ray,JournalofAppliedPolymerScience,2004,93,122-130.
169. H.L.Chou,C.M.Wu,F.D.LinandJ.Rick,AIPAdvances,2014,4,087111.
170. P.K.Khanna,R.Gokhale,V.V.V.S.Subbarao,A.K.Vishwanath,B.K.DasandC.V.V.
Satyanarayana,MaterialsChemistryandPhysics,2005,92,229-233.
171. G.MpourmpakisandD.G.Vlachos,PhysRevLett,2009,102,155505.
172. R.Patakfalvi,Z.VirányiandI.Dékány,ColloidPolymSci,2004,283,299-305.
173. L.Zhao,D.Jiang,Y.Cai,X.Ji,R.XieandW.Yang,Nanoscale,2012,4,5071-5076.
174. R.Baber,L.Mazzei,N.T.ThanhandA.Gavriilidis,JournalofFlowChemistry,2016,1-11.
175. D.Liu,S.Cito,Y.Zhang,C.F.Wang,T.M.SikanenandH.A.Santos,AdvancedMaterials,2015,
27,2298-2304.
176. R.Couto,S.Chambon,C.Aymonier,E.Mignard,B.Pavageau,A.ErriguibleandS.Marre,Chem.
Commun.,2015,51,1008-1011.
177. T.Gendrineau,S.Marre,M.Vaultier,M.PucheaultandC.Aymonier,AngewandteChemie
InternationalEdition,2012,51,8525-8528.
178. B.Giroire,S.Marre,A.Garcia,T.CardinalandC.Aymonier,ReactionChemistry&Engineering,
2016,DOI:10.1039/C5RE00039D.
179. N.Acharya,M.SenandC.Hsueh-Chia,IntJHeatMassTran,1992,35,2475-2489.
180. N.Acharya,M.SenandH.-C.Chang,IntJHeatMassTran,2001,44,3189-3199.
181. M.M.Mandal,C.Serra,Y.HoarauandK.Nigam,MicrofluidNanofluid,2011,10,415-423.
182. A.K.SaxenaandK.D.P.Nigam,AIChEJournal,1984,30,363-368.
183. J.Reijenga,G.Aben,T.P.VerheggenandF.Everaerts,JChromatogrA,1983,260,241-254.
184. P.K.Jain,K.S.Lee,I.H.El-SayedandM.A.El-Sayed,TheJournalofPhysicalChemistryB,2006,
110,7238-7248.
185. X.Liu,M.Atwater,J.WangandQ.Huo,ColloidsandSurfacesB:Biointerfaces,2007,58,3-7.
186. T.Hendel,M.Wuithschick,F.Kettemann,A.Birnbaum,K.RademannandJ.Polte,Analytical
Chemistry,2014,86,11115-11124.
187. M.B.Hariri,A.DolatiandR.S.Moakhar,JournalofTheElectrochemicalSociety,2013,160,
D279-D288.
188. L.DjenidiandB.Moghtaderi,JournalofFluidMechanics,2006,568,223-242.
160
189. W.G.Burns,M.MatsudaandH.E.Sims,JournaloftheChemicalSociety,FaradayTransactions,
1990,86,1443-1447.