Microfluidics at high-intensity X-ray sources: from microflow chips to microfluidic liquid jet systems

Mic o luidics a high-in ensi y X- ay sou ces: om mic o low chips o mic o luidic liquid je sys ems DOCTORAL THESIS (Disse a ion) o be awa ded he deg ee o Doc o e um na u alium (D . e . na .) a he Facul y o Biology, Chemis y and Ea h Sciences, Uni e si y o Bay eu h submi ed by Dipl. Chem. Ma in T ebbin om Hambu g Bay eu h, 2013 The wo k desc ibed in his hesis was ca ied ou a he Ins i u e o Physical Chemis y a he Uni e si y o Hambu g (Augus 2009 o Sep embe 2010) and he Depa men o Physical Chemis y I a he Uni e si y o Bay eu h (Oc obe 2010 o Sep embe 2013) unde he supe ision o P o . S ephan Fö s e . This is a ull ep in o he disse a ion submi ed o ob ain he academic deg ee Doc o o Na u al Sciences (D . e . na .) and app o ed by he Facul y o Biology, Chemis y and Geosciences o he Uni e si y o Bay eu h. Ac ing dean: P o . D . Rhe Kempe Thesis submi ed: 13.09.2013 Da e o Scien i ic Colloquium: 19.12.2013 Doc o al Commi ee: P o . D . S ephan Fö s e (1s e iewe ) P o . D . And eas Fe y (2nd e iewe ) P o . D . Hans-We ne Schmid (Chai man) P o . D . Thomas Scheibel "I you wan be e esul s, you ha e o ask you sel be e ques ions.” - unknown Table o con en s Summa y…………………………………………………………….….….………….13 Zusammen assung ………………………………….……………………………. 15 1 In oduc ion …………………………………………………...…………………….17 1.1 Mo i a ion ………………………………………………………………………….17 1.2 Mic o ocus X- ay sou ces ……………………………………………………....17 1.3 In oduc ion o mic o luidics …………………………….……………………. 20 2 Mic o luidic de ices as expe imen al pla o ms o X- ay s udies ….. 25 2.1 In oduc ion ………………………………………………………………………. 25 2.2 X- ay compa ible mic o luidic de ice ypes ……………………………..…. 29 2.3 Ou look: new p ocessing echniques and de ice ma e ials ……………... 34 2.4 Mic o luidic liquid je de ices o nex gene a ion X- ay sou ces ….…. 35 2.5 Conclusions ……………………………………………………………………......39 3 Theo e ical Fundamen als ………………………………………………… 41 3.1 Sel -o ganiza ion o amphiphiles ……………………………………….….…. 41 3.2 Fluid dynamics undamen als o mic o luidics …………………………... 43 3.2.1 Na ie -S okes equa ions ……………………………………………….…… 43 3.2.2 The No-Slip condi ion ……………………………………………………… 45 3.2.3 Con ec ion and Di usion …………………………………………………... 46 3.3 Solu ion o non-linea p oblems ………………………………………………48 3.3.1 Fluid s uc u e in e ac ion ………………………………….……………….48 3.3.2 Shea o ces and Non-New onian luids ………………………….….….… 48 3.3.3 Two-phase sys ems and liquid je s …………………………………………. 49 3.4 Small angle X- ay sca e ing ………………………………………………….. 53 3.4.1 P e ace ……………………………………….….…….….….……………… 53 3.4.2 In oduc ion ………………………………………………………………… 53 3.4.3 The sca e ing ec o …………………………………………………..….... 54 3.4.4 The sca e ing pa e n ………………………………………………………. 55 3.4.5 Fo m ac o …………………………………………………….……………. 58 3.4.6 S uc u e ac o ……………………………………………………………... 59 3.4.7 B agg e lexes ……………………………………………………………….. 60 3.4.9 Model-based o de analysis ………………………………………………… 63 3.4.10 Pa icle o ien a ion dis ibu ion …………………………………………... 64 4 Me hods and Techniques ………………………………………………….. 71 4.1 Pho oli hog aphy ………………………………………………………………… 71 4.2 Molding Ma e ials ………………………………………………………………. 75 5 Summa y and S a egy …………………………………………………….. 79 quan i izie . Die S ömungso ien ie ung on zylind ischen, Wu m- und Fib illen-a igen S uk u en is on zen ale Bedeu ung ü die iele P ozesse, wie z. B. de Fase he s ellung, den Sp i zguss-Ve ah en ode dem Fluss on Zellen und P o einen du ch dünne Kapilla en. Ein wei e es Beispiel ü geschlossene Kanäle demons ie die hohe P obene izienz mik o luidische Rillen-Chips. Diese sind nu wenige Millime e g oß, sche -induzie end und benö igen lediglich 2-5 µl P obe ü die Rön genanalyse eines sche -o ien ie e n Polyme - Nanokomposi Ma e ials. De zwei e Typ on Mik o luidik-Chips basie au eine o enen Düsen-Geome ie und p oduzie Flüssigkei s-S ahlen mi Mik ome e -D uchmesse n (0.9 bis 5 µm) bei seh ge ingen Fluss a en (150 o 1000 µl h-1) un e No mald uck- ode Vakuum-Bedingungen. Die o ges ell en mik o luidischen Flüssigkei s-S ahl-Chips basie en au dem P inzip de gas-dynamischen i uellen Düse (GDVN), welches de zu e lässigen und e s op ungs eien Be ieb übe lange Zei en e möglich . Da übe hinaus sind diese Chips ein ach und mi els so li hog aphische Techniken he s ellba , welches ein p äzises und ep oduzie ba es Mik okanal-Design e möglich , dass k i isch ü die Op imie ung on Flüssigkei s-S ahlen bei kleinen Fluss a en is . Diese Design-Kon olle wi d demons ie du ch die ein ache In eg a ion zusä zliche mik o luidische Elemen e, wie z. B. ein S ahl- im-S ahl-S ömungs okussie ung ode dich e G uppen iele Düsen, ohne zusä zliche He s ellungssch i e. Das mik o luidische Flüssigkkei s-S ahl Sys em wu de eben alls mi els eine Reihe mik oskopische Me hoden un e such (s. oben) sowie CFD- Simula ionen. Zusammen mi de Va ia ion expe imen elle Pa ame e sowie Kanalgeome ien, haben diese Un e suchungen zu einem besse en Ve s ändnis des luiddynamsichen Ve hal ens on Flüssigkkei ss ahlen in mik o luidischen Chips sowie zu de Kon olle des S ahld uchmesse s und des T op enabb uch-Typs ge üh . Diese e wähn en Vo eile (Zu e lässigkei , ge inge P oben e bauch, e c.) und die o ene Kanalgeome ie machen diese mik o luidischen Flüssigkei s-S ahl Sys eme höchs ele an ü die E ablie ung als P obenumgebung an Rön gen-FEL. Diese Ein ich ungen gene ie en Rön genbli ze, die ul aku z ( s-Be eich) und eno m in ensi sind und mi denen ein ganzes S eubild mi nu einem Lich puls au genommen we den kann. In diesem P ozess explodie die P obe und wi d in ein leuch endes, ca. 60 000 K-heißes Plasma e wandel . Deshalb sind s a ische P oben ode geschlossene Flusssys eme die diesen Rön genquellen de nächs en Gene a ion inkompa ibel, was die Wich igkei des Ansa zes o ene Flüssigkei s-S ahlen un e s eich . De wissenscha liche Wei blick des mik o luidischen Konzep s und de li hog aphischen Mik o ab ika ion wu de eben alls demons ie , indem zusä zlich Lösungen ü ande e Rön gen- eie Anwendungen ge unden wu den. Die Beispiele hie ü um assen CFD- simula ionen ü nich -linea e wissenscha liche P oblems ellungen eines Sp üh ockne s, genaue dessen in e ne Fluid-S uk u -In e ak ion, ode die Ges al ung & He s ellung on Mik os uk u en ü den Mik o-Kon ak d uck, also das mik ome e -genaue S empeln, kugel ö mige Polyelek oly -Bü s en. Ein wei e es Beispiel demons ie die Kombina ion on Nano echnologie und Mik os uk u ie ungs-Techniken ü En wicklungen in Rich ung senso ische Anwendungen. 16 1 In oduc ion 1.1 Mo i a ion A wide ange o na u e’s undamen al g ow h p inciples, like o example sel -assembly o nuclea ion & g ow h, a e go e ned by p ocesses on he nanome e o mic ome e leng h scale and hei unde s anding is o g ea in e es o mode n ma e ial sciences and condensed ma e physics. 1-3 The in es iga ion o hese elemen al p inciples equi es he combina ion o ad anced analysis echniques ha ex end in o he ele an leng h scales wi h sample en i onmen s ha a e capable o con olling he expe imen al physical and chemical condi ions wi h e y high p ecision. F om an analysis poin o iew, X- ay sca e ing echniques and mic oscopic me hods o e a a ie y o ools o s udy hese p ocesses wi h g ea de ail on all he ele an leng h scales. Wi h he cu en de elopmen s in synch o on echnology and wi h he ad en o ee elec on lase s (FEL), e y exci ing possibili ies a ise such as ime- esol ed expe imen s a highly b illian synch o on sou ces and se ial em osecond X- ay nanoc ys allog aphy a X- ay ee elec on lase s. 4-7 Wi h espec o he sample en i onmen , he in e disciplina y scien i ic ield o mic o luidics is p edes ined o undamen al in es iga ions because i ocuses on he p ecise con ol and manipula ion o e y small luid olumes in mic ome e -sized geome ies. 8 The combina ion o mic obeam X- ay sca e ing, mic oscopy and mic o luidics is cu en ly being de eloped in o a powe ul expe imen al me hodology sui able o he in si u in es iga ion o nanos uc u es, pa icle alignmen and he in si u s udy o kine ics. 9-14 This p og ess is enabled by c ea ing X- ay compa ible mic o low chips and mic o luidic liquid je de ices ha allow new scien i ic app oaches compa ed o adi ional expe imen s on he mac oscale because he sample’s en i onmen al condi ions, like small scale concen a ion g adien s o low ields on he mic on-scale, can be ailo ed o pa icula scien i ic ques ions. 9,13,15 The ollowing chap e s will p o ide an o e iew o e his combina ion o echnologies by co e ing i s o igins & challenges, he heo e ical undamen als as well as he in ol ed me hods and echniques. 1.2 Mic o ocus X- ay sou ces Small-angle X- ay sca e ing (SAXS) allows o analyze he nanome e -sized in e nal s uc u e o a sample. Howe e , X- ay sou ces wi h adi ional collima ion sys ems a e only capable o deli e beams which a e ypically millime e -sized. This ange o X- ay spo sizes p e en s he p ecise analysis o small o he e ogeneous samples due o he signal a e aging o e he illumina ed a ea. 17 This si ua ion imp o ed emendously wi h he g ea ly inc eased peak b illiance o s a e-o - he-a synch o ons and wi h he ad en o specialized echniques o he gene a ion o mic o ocused X- ays. 4,16 Due o he pionee ing wo k a dedica ed ocused X- ay beamlines, a a ie y o ocusing p inciples has been de eloped ha in ol e componen s like Ki kpa ick- Baez (KB) c ossed mi o s o -mul ilaye s, capilla ies, F esnel op ics, wa e guides o compound e ac i e lenses (CRL) and oge he wi h ad ances in high-p ecision posi ioning sys ems, hese echniques now enable X- ay oci down o he nanome e ange. 11,16-26 Consequen ly, samples can now be analyzed wi h a much highe spa ial esolu ion and g ea e p ecision. This capabili y allows o dis inguish be ween di e en ly s uc u ed domains in a ma e ial o which he sca e ing signal p e iously had been a e aged. Howe e , he ocus spo size is no he only hing o conside o expe imen s wi h ocused X- ay beams. I is also e y impo an o conside he di e gence o he beam as well as he o e all pho on lux. 20 The ela ion be ween hese h ee pa ame e s can be imagined as a iangle because all o hem a e in e ela ed. Fo example, i is possible o gene a e e y small- sized X- ay beams using CRL- ocusing op ics. This also inc eases he pho on lux a he ocus posi ion emendously. Howe e , his is only possible a he cos o a highe beam di e gence and consequen ly a loss o esolu ion. I is also possible o ha e a small beam wi h low di e gence o an inc eased esolu ion, bu his migh only be possible a he cos o pho on lux which esul s in longe measu emen imes. I pho on lux and di e gence a e bo h impo an , he minimum X- ay spo size has o be la ge wi h a dec ease o spa ial esolu ion, al hough he beam size would s ill be in he low mic ome e ange. An example o e iew o e he spo sizes and i s in luence on di e gence and lux o di e en ocusing echniques is shown in Fig. 1. Figu e 1 Compa ison o di e en X- ay mic o ocusing echniques and hei spo size- dependen in luence on lux and di e gence. The ocusing ypes include capilla ies ( op), Ki kpa ick-Baez (KB) c ossed mi o s (middle) and compound e ac i e lenses (CRL, bo om). (Figu e om 20, Copy igh IOP Publishing) 18 Today’s dedica ed mic o ocus X- ay beamlines a 3 d gene a ion synch o ons, like o example P03 (PETRA III a DESY, Hambu g, Ge many) o ID13 (ESRF, G enoble, F ance), also ake he ocusing dis ance om he op ics o he sample in o accoun . This leads o a lowe di e gence a a gi en ocus spo size while a high pho on lux is main ained. In he end, i is a ques ion o he op imum combina ion o ocusing se ings o a gi en sample o expe imen al se up. In case o a mic o luidic X- ay expe imen , he maximum di e gence o he mic o ocused beam is dic a ed by he channel heigh which ypically anges be ween 50 µm and 100 µm. In mos o ou mic o luidic SAXS expe imen s, he X- ay beam size was adjus ed o spo sizes a ound 10 µm by 10 µm o gua an ee a low di e gence and main ain a high pho on lux. An example which clea ly demons a es he bene i s o he high spa ial esolu ion o mic o ocused X- ay beams is he analysis o hin cellulose- o high-pe o mance polyme ibe s27-29 A ibe ’s s uc u e can be mapped in g ea de ail which also p o ides in o ma ion abou he ibe s in e nal s uc u e. Ano he e y in e es ing example o a biological sample sys em a e spide silk ibe s. 22,30-34 These in i o and in i o s udies e ealed how he spide silk ibe ’s nanos uc u e changes du ing he spinning p ocess, du ing i s elonga ion o unde he in luence o and how his a ec s he ibe ’s ensile s eng h and mic o-s uc u al p ope ies. The smalle X- ay beam size is also impo an o g azing-incidence small-angle sca e ing (GISAXS) because he mic o ocus-illumina ed a ea is much smalle compa ed o con en ional X- ay beams, enabling be e spa ial scanning esolu ion. 35 Ano he example whe e he use o a mic o ocused X- ay beam g ea ly imp o es he spa ial esolu ion is SAXS mic o omog aphy. 36-40 This measu emen echnique e eals he in e nal h ee-dimensional s uc u e o a sample, like o example a high pe o mance polyme ibe , by o a ing i du ing he de ailed mapping wi h he X- ay beam. Due o he Nyquis -Shannon sampling heo em and he mic o ocused X- ays ha enable he mapping o much smalle olumes, he sample scans equi e much less images (and he e o e sho e scanning imes) o a gi en esolu ion compa ed o omog aphic scans wi h la ge beams. 41,42 The ongoing de elopmen o X- ay sou ces leads o inc easingly b illian and in ense beams. 4 This can lead o new p oblems conce ning he sample: he maximum X- ay dose be o e he sample deg ades. Biological samples o so ma e a e jus wo examples o sample sys ems which a e suscep ible o adia ion damage. 43 A p o ein c ys allog aphic case s udy demons a ed ha i is possible o inc ease he sample’s esilience o beam damages h ough eezing. 44 Ob iously, his c yogenic app oach is imp ac ical o liquid samples o solu ions ha eeze below he a ge ed empe a u e, bu i is a e y use ul me hod o p o ein c ys als. Howe e , his s udy also e ealed ha he success ul collec ion o a c ys allog aphic da a se is only possible up o a ce ain X- ay dose because o he wise he sample deg ades be o e he da a se is ob ained. 44 This maximum dose dic a es a minimum c ys al size in he mic ome e 19 ange which is e y limi ing o he p o ein-s uc u e de e mina ion a adi ional X- ay sou ces like synch o ons o lab sou ces. The equi ed minimal c ys al size could be educed by he use o mic o ocused X- ays in combina ion wi h mo o ized sample handling ha enable p ecise c ys al scans, bu i is s ill an ex emely challenging ask o g ow ’la ge’ mic ome e - sized p o ein single c ys als in he i s place. 43,45 One possible app oach o a oid he beam damage o liquid samples o solu ions is o use a con inuous sample s eam minimizing he sample esidence ime in he olume o he X- ay ocus. This app oach equi es a sample en i onmen ha allows o con ol liquids wi h e y high p ecision and ep oducibili y on he mic ome e scale. I his was no he case, he possibly high sample consump ion a es o mac oscopic low sys ems would make he equi ed sample amoun s imp ac ical. This is whe e he combina ion o mic o ocused X- ay beams and X- ay compa ible mic o luidic de ices becomes impo an and demons a es i s s eng hs. 9,46-48 The challenges and bene i s o X- ay compa ible mic o luidic sample en i onmen s as well as i s examples o expe imen s will be discussed in u he de ail in he ollowing chap e . Ano he app oach o o e coming he maximum dose limi o solid and liquid samples becomes a ailable wi h he ad en o X- ay ee elec on lase s (XFEL). 49-51 This nex gene a ion o X- ay sou ces sol es he adia ion damage p oblem by gene a ing highly in ense em osecond X- ay pulses and applying he p inciple o ‘di ac ion be o e des uc ion’. 6,52 Each gene a ed X- ay pulse is so in ense ha a ull di ac ion pa e n is collec ed wi h his single sho while i s pulse leng h is so sho ( em osecond ange) ha he di ac ion pa e n is gene a ed be o e he sample is des oyed by Coulomb explosion. Consequen ly, he success ul collec ion o di ac ion pa e ns equi es a esh sample wi h each single pulse. Howe e , he ealiza ion o an e icien and eliable way o gene a ing a con inuously eplenishing sample s eam is a e y challenging ask. 53,54 This is whe e he mic o luidic liquid je de ices, which a e in oduced in mo e de ailed in chap e 7.3, o e g ea po en ial as a sample en i onmen o XFELs and o he pulsed lase expe imen s due o he lexible design con ol and as ab ica ion ou ines o hese de ices. 1.3 In oduc ion o mic o luidics Mic o luidics has become oday’s pla o m echnology o he p ecise con ol and manipula ion o luids on he mic ome e scale. 8 The e m ‘mic o luidics’ o igina es om he combina ion o mic o ab ica ion echniques such as so li hog aphy wi h he luid dynamics on he mic ome e scale. 10,55-57 Acco dingly, small amoun s o luids ( o li e s) a e handled and manipula ed using channels wi h dimensions o ens o hund eds o mic ome e s. 58 Hence, he bigges ad an age o mic o luidics lies in he mic ome e dimensions o he channels and he ela ed luid dynamic implica ions o he samples such as lamina low o di usion-based mixing. These undamen al physical p ope ies a he mic on scale enabled he e olu ion o a a ie y o mic o luidic ools. As an example, hese ools can be highly 20 bene icial o he p ecise s udy o nuclea ion and g ow h p ocesses and he as and e icien sc eening o expe imen al condi ions, like i.e. pH, ionic s eng h, species composi ions, shea o ces, cosol en s and concen a ion. 58 In he beginnings o his echnology, he mic o ab ica ion o hese small channels has been s ongly in luenced by he ield o mic oelec omechanical sys ems (MEMS) ha in ol es elec onic ci cui s, senso s and mic omechanical componen s. 58-61 This ield o e s a ich po olio o a ailable glass- and silicon- ela ed ab ica ion echniques which s imula ed he ea ly de elopmen o mic o luidics. 58 Today, he mos used mic o luidic de ice ab ica ion echnique is so li hog aphy which is he combina ion o so ma e ials such as polydime hylsiloxane (PDMS) wi h pho oli hog aphy. 55,56,62,63 These ab ica ion me hods and mo e de ice ma e ials will be discussed in g ea e de ail in he chap e s 2 and 4. The closely ela ed e ms ‘minia u ized o al analysis sys ems’ (µTAS) o ‘lab on a chip’ o igina e om he 1990s. 64 They desc ibe he concep o combining he elemen s o mic ochemical ‘ ac o ies’ on a small, single chip which can inco po a e unc ional elemen s such as pumps, al es, mixe s, swi ches, hea e s, mul iplexe s, elec odes and senso s. 8,58,65-83 Consequen ly, he ‘lab on a chip’ concep aims owa ds he inc ease o mobili y and he educ ion o ene gy consump ion, was e p oduc ion and ul ima ely p oduc ion cos s by elimina ing he need o adi ional labo a o y equipmen . 64 This idea has been demons a ed o complica ed chemical eac ions and complex mic ochannel ne wo ks ha combine mul iple unc ional elemen s on a single chip. 67,79 The concep o a mic o eac ion plan on a chip has also been demons a ed o complex eac ions like he li ing anionic polyme iza ion o block copolyme s wi h di ec on-chip DLS analysis o he esul ing micelles. Ano he complica ed eac ion on a chip has been demons a ed o he syn hesis o 18F-labeled o ganic compounds ha a e used in posi on emission omog aphy (PET). 58,81,84,85 Addi ionally o he al eady-men ioned ea u es o mic o ludics ha include low sample consump ion, he bene icial ea u es o mic o luidics also include he in eg a ion o unc ional elemen s on a chip which enable small de ice oo p in s. Fu he , he ab ica ion cos s a e ypically small, he was e p oduc ion is minimized and i is also possible o un exo he mic eac ions while main aining empe a u e con ol. 86,87 This g ea empe a u e con ol is enabled by o he small amoun s o eac ing mass combined wi h he high su ace o olume a io o he mic ochannel ne wo k. As a consequence, sa e ope a ion is gua an eed while he uni o m hea ans e also gi es g ea con ol o e he eac ion kine ics. 86,87 Fu he examples include he p oduc ion o mic opa icles and nanopa icles wi h a la ge di e si y o mo phologies and physicochemical p ope ies wi h espec o size, shape, su ace cha ge and amphilici y. 88-96. Al hough he olumes o he handled luids a e ypically small, he massi e pa alleliza ion o mic o luidic de ices o e s he po en ial o upscaling he p ocesses o indus ial scales. 97 21 Mic o luidic echnology also o e s many ad an ages when i comes o sample analysis and, consequen ly, oday’s lis o de eloped applica ions o mic o luidic pla o ms is mani old. 58 Fo example, sample analysis ela ed demons a ions include “sepa a ions coupled o mass spec oscopy, high- h oughpu sc eening in d ug de elopmen , bioanalyses, examina ion and manipula ion o samples consis ing o a single cell o a single molecule“.58,74,96,98-101 Fu he , applica ions include p ocesses such as ee- low elec opho esis o blood sample analysis which ha e been imp o ed and minia u ized. 102,103 Toge he wi h he abo e-desc ibed unc ional elemen s, like al es and pumps, combina o ic expe imen s and high- h oughpu eac ion sc eenings became possible 67,79,104,105 As an example, Quake e al. de eloped me hods o he mic o luidic la ge scale in eg a ion which is he mic o luidic analogue o he echnological jump om single ansi o s o mic op ocesso s in elec ical enginee ing. 67,104 Th ough mic o luidic al es, pumps, and mul iplexe s, his echnology enables combina o ics and high- h oughpu sc eenings (HTS) o single cell analysis, deoxy ibonucleic acid (DNA) syn hesis, digi al polyme ase chain eac ion (dPCR), genome sequencing, as well as and la ge scale genomics and p o eomics. 80,106-117 Fu he mo e, his HTS-app oach also allows o ind and op imize p o ein c ys alliza ion condi ions while only equi ing e y small amoun s o sample. 118-121 Figu e 2 Illus a ion o he mic o luidic la ge scale in eg a ion concep . 67,79,104 (A,B) The low and mixing o samples wi h nanoli e olumes a e con olled using small al es and pumps. Acco dingly, he esul ing de ice oo p in is e y small (C) and he compu e -con olled de ices can handy e y complex asks such as he high- h oughpu sc eening o luo escence- based single-cell assays (D). (Images om 67(A,B, Copy igh Science), 79 (C, D, Copy igh Na u e)) 22 Ano he example o mic o luidic condi ion sc eening and combina o ics is he c ys alliza ion o single p o ein c ys als. A wide ange o mic o luidic ools ha e been de eloped o his pu pose due o his ield’s g ea impo ance o medicine and he li e sciences. 119,120,122-128 These examples show ha i is possible o gene a e sc eening lib a ies o he au oma ed c ys alliza ion o enzymes, p o eins and o he subs ances unde de ined condi ions in d ople s o indi idually add essable, on-chip mic ocompa men s while only equi ing e y small amoun s o sample. All hese examples demons a e ha mic o luidics o e s g ea con ol o e luids, eac ions and expe imen al condi ions by aking ad an age o he luid dynamics on he mic ome e scale ha enable lamina low and di usi e mixing. 10 In combina ion wi h he abo e- men ioned mic o ocus X- ay echniques, many new expe imen al oppo uni ies a ise which would no be possible wi h con en ional mac oscopic sys ems. Howe e , he ans e o mic o luidic echnology o X- ay expe imen s is echnically e y challenging due o he X- ay compa ibili y o he di e en de ice ma e ials. In his ongoing ans e p ocess, a a ie y o ab ica ion app oaches and mul iple de ice ypes ha e been de eloped. The nex chap e will e iew hese a ailable X- ay compa ible de ice ypes and desc ibe he s udies ha ha e been pe o med a mic o ocus X- ay sou ces. 23 24 2 Mic o luidic de ices as expe imen al pla o ms o X- ay s udies 2.1 In oduc ion The combina ion o mic o luidics and mic o ocused X- ay beams is a ela i ely young ield o esea ch ha s a ed wi h ea ly expe imen s by Pollack e al. du ing he up ise o mic o ocused X- ays. 46-48 These sample en i onmen s ha e been used o kine ic olding s udies o biological sys ems such as RNA o p o eins. 46-48 Hence, his echnology-combining app oach links he bene i s o small X- ay mic obeams wi h he ones om mic o luidics and, he e o e, allows o s udy he sample sys em’s esponse in si u and in ope ando wi h high p ecision and con ol. 9-12 The esul ing oppo uni ies as well as he i s app oaches and cu en de elopmen s o mic o luidic sample en i onmen s o X- ay expe imen s will be e iewed o e he cou se o his chap e . As men ioned abo e, he u bulence- ee low condi ions wi hin he mic ochannels o e g ea con ol o e he chemical and physical condi ions. These expe imen al pa ame e s, like concen a ion, pH, shea - and ex ensional o ces, can be con olled p ecisely and ep oducibly by adjus ing low, mixing and he mic ochannel geome y acco ding o he sample sys em o in e es . Fo example, eac ion kine ics a e di usion-based due o he lamina low. These luid dynamic condi ions enable de ailed scans o he mixing a ea wi h mic o ocused X- ay beams and esul s in obse able eac ion ime scales ha ex end o e ange om ze o o se e al seconds. 46-48,129-132 Fu he , he ea lie -men ioned p oblem o adia ion damage (chap e 1.2) can be ackled by using sample en i onmen s ha eplenish he sample con inuously and he e o e educe he X- ay dose o a oid beam damage o he sample. This is pa icula ly impo an o as - deg ading sys ems, like o example biological samples. While he dose ba ie o hese samples is abou 200 pho ons o X- ays a an ene gy o 12 keV, he X- ay dose in a ypical mic o luidic con inuous low expe imen is educed o doses on he o de o 0.1 pho ons due o he e y sho sample esidence imes in he exposed olume. 9,44,49 The expe imen al se up can be op imized e en u he by uning he X- ay ene gy o maximize he ansmission h ough he de ice’s mic ochannel ma e ial and he liquid he ein. 133 Nex o he capabili y o success ully measu ing as -deg ading and X- ay sensi i e samples, mic o luidic de ices enable he s udy o p ecious samples ha a e only a ailable in small amoun s, like i.e. memb ane p o eins o deoxy ibonucleic acid (DNA). 46-48,130-132 T adi ional expe imen s could p e en hei analysis due o he limi ed sample a ailabili y and he expe imen ’s minimum equi ed sample quan i y while mic o luidic de ices a e highly sample e icien and consume only e y small olumes in he ange o mic o- o e en nanoli e s pe hou . 67,79,118,134 25 Since he so li hog aphic eplica ion using polydime hylsiloxane (PDMS) is so easy o lea n, i has become one o oday’s mos used ab ica ion echnique o mic o luidic de ices. 58 Du ing he mic os uc u e eplica ion he cu ing PDMS sh inks less han 1% which esul s in e y p ecise eplicas ha can be used o example as s amps in mic o con ac p in ing. 151-154 These open eplicas can also be sealed o c ea e closed mic ochannels. This is ypically achie ed by ac i a ing he PDMS using ai plasma and hen binding i co alen ly o glass slides which esul s in e y p essu e esis ance mic ochannels. 57 This combina ion o PDMS ha is bonded glass is ypically he ma e ial combina ion o choice when i comes o mic oscopy- ela ed applica ions o mic o luidics due o i s excellen op ical anspa ency om 240 nm o 1000 nm and i s low oxici y. 58 Addi ionally, he elas ici y o PDMS can be con olled by he a io be ween oligome he c oss-linke while his ma e ial’s gas pe meabili y is bene icial o cell cul u es o mic oe apo a ion. 9,15,58 PDMS also has wo signi ican d awbacks: he limi ed compa ibili y wi h sol en s and he de ice ouling om he unspeci ic adso p ion o biomolecules. 9,34,147 Howe e , PDMS o e s a wide ange o su ace modi ica ion possibili ies which enables o minimize hese d awbacks o a oid he nega i e e ec s all oge he . 155-159 The modi ica ion ou ines include sol-gel glass coa ings o imp o ed sol en esis ance, laye -by-laye deposi ion o polyelec oly es o pe manen ly hyd ophilic channel walls, he co alen deposi ion o luo ina ed epellan s agains de ice ouling and UV-con olled pho ochemis y o g a ing eac ions on he channel su ace o we abili y uning. 155-159 PDMS can also be used o he measu emen o SAXS di ec ly i he de ices a e e y hin. This has been demons a ed a he Diamond ligh sou ce in o he o ien a ion analysis o lamellae in mic ochannels. 160 Thin PDMS-based de ices also allow mic oe apo a ion ha can be used o he gene a ion o concen a ion g adien s in non- lowing samples. 15 Al hough showing a s onge backg ound signal, hese de ices’ ma e ial homogenei y ypically allows o ex ac he sample’s signal h ough backg ound sub ac ion. Howe e , his p ocess can dec ease he signal quali y and e en ually he abili y o e alua e he ob ained sca e ing pa e ns. This dec eased signal- o-noise a io o he sample can esul om ac o s like he subs ac ion-induced educ ion o he signal and added de ec o eadou noise om combining mul iple iles. 161 Addi ionally o he X- ay backg ound signal, he sol en compa ibili y o his ma e ial is also e y limi ed. 147 A much be e sol en compa ibili y is p o ided by glass-only-based mic o luidic de ices. As an example, glass capilla y de ices wi h ube-in- ube geome ies ha e been success ully ope a ed a synch o ons o he s udy o in si u spide silk ibe o ma ion. 22 This de ice ype also o e s he bene i o educed o no wall clogging due o he coaxial liquid shea h o he ou e capilla y. 162 Fu he , he glass capilla ies can also be e ched down o hicknesses a ound 50 µm o imp o ed X- ay ansmission. The main downside o his de ice ype is i s ab ica ion p ocedu e because i is complex, in ol es p ecise manual capilla y alignmen skills and las ly, he design a ia ions o he channel geome y a e limi ed. 22,162 32 When i comes o X- ay ansmission expe imen s, glass and PDMS o hei combina ion is coun e p oduc i e because o he ma e ial’s own s ong small angle sca e ing signal. 163 Al e na i ely, mic o luidic de ices can be di ec ly ab ica ed hough lase abla ion. 141 The mic os uc u es a e w i en di ec ly in o Kap on ilm which is hen sealed wi h ano he Kap on ilm o yield he closed mic ochannel. The esul ing de ices a e e y hin, sol en esis an and hey show good mechanical p ope ies. Fu he , Kap on is an excellen ma e ial o X- ay applica ions because o i s high adia ion esis ance and low abso p ion. I has also been shown ha he lase abla ion app oach can be applied o o he ma e ials such as poly(me hyl me hac yla e) (PMMA), polys y ene (PS) and cyclic ole in copolyme s (COC). 164 Howe e , he majo d awback o his lase -based app oach is he low machining speed which can lead o long p ocessing imes due o he ypically wide-sp eading mic ochannel pa e ns. 141 Ano he a ian o s able mic o luidic de ices wi h Kap on as he only window ma e ial ha e been desc ibed by P ohl e al. 23,130,139,140 In hese examples, open mic ochannels a e spa k e oded in o s ainless s eel pla es which a e hen sealed a he op and bo om using sel - adhesi e Kap on ilm. The minimum channel size is es ic ed o abou 60-100 µm due o he esolu ion o he spa k e oding echnique. 9 While his minimum channel size can be su icien o many mic o luidic expe imen s, he main d awback o his echnique is he limi ed design lexibili y because i is only possible o c ea e ela i ely simple s uc u es like s aigh lines. 9 This design lexibili y can be inc eased by choosing a channel wall ma e ial ha is sui able o apid p o o yping. 57 A widely used p ocess is called so li hog aphy and i in ol es mic ochannel empla es ha a e ab ica ed by mic os uc u ing a pho o esis on a silicon wa e using UV li hog aphy. 56 A hin laye o moldable ma e ial, such as PDMS, is hen cas ed on his empla e by doc o blading o spin coa ing o gene a e open mic ochannels. These channels a e hen sealed wi h Kap on ilms om op and bo om, simila o he s eel- based de ices ha ha e been desc ibed abo e. 13,130,131,139,165 This combina ion o e s g ea design lexibili y h ough apid p o o yping while main aining good X- ay p ope ies by using Kap on windows. Howe e , he X- ay signal could be in luenced by he adhesi e laye o he sealing kap on ape and he use o PDMS also limi s he sol en compa ibili y o hese de ices. 147 A di e en ou ine ha akes ad an age o moldable ma e ials and apid p o o yping is he ab ica ion o mic o luidic de ices made o an UV-cu able op ical adhesi e by No dland (NOA81). 14,132,166,167 O iginally used o he glueing o op ical componen s such as lenses, his hiol-ene-based ma e ial is cu ed by a adical mechanism ha allows su ace chemis y modi ica ion and which is ini ia ed by a UV-sensi i e ini ia o . 149,168 This ma e ial also allows o ab ica e e y hin de ices wi h window hicknesses o a ew ens o hund eds o mic ons while he small angle backg ound signal is also much lowe compa ed o PDMS. 9,132,135 Fu he mo e, i p o ides a much be e esis ance o a wide ange o sol en s compa ed o 33 PDMS. 147,149 I has also been demons a ed ha his ma e ial is sui able o he ab ica ion o mic o luidic h ee-dimensional low geome ies by aligning and sealing wo mic os ucu ed laye s. 132 Each side is ab ica ed by mul ilaye li hog aphy which yields a low ocusing geome y ha minimizes o p e en s clogging o he sample o he mic ochannel wall. 9,132,162 Howe e , he exac alignmen o he wo hal es wi h mic ome e p ecision can be e y challenging. A gene al downside o his ma e ial is he obse able beam damage o his ma e ial, especially o highe X- ay in ensi ies, which can be obse ed as a b ownish spo on he yellow-whi e anslucen ma e ial. Howe e , his beam damage does no in e e e wi h he lowing samples on he imescale o ypical mic o luidic expe imen s. 9,132 2.3 Ou look: new p ocessing echniques and de ice ma e ials New ways o ab ica e mic o luidic de ices a ise wi h he ad en o new ma e ials, new p ocessing echniques o he ecombina ion o exis ing componen s. As men ioned ea lie , so a no ma e ial o ou ine has been s udied ye ha ma ches all c i e ions simul aneously, bu many good ou ines ha e been desc ibe o he speci ic expe imen s. Ma e ials which allow apid p o o yping, like PDMS o NOA81, can ake ad an age he g ea design lexibili y ha is enabled h ough so li hog aphy. 57 O he igid o ha d ma e ials, in e ms o ma e ial p ope ies o ease o p ocessing (i.e. silicon, glass, Kap on), usually ely on ab ica ion p ocesses ha can be complica ed o do no allow he same design lexibili y as so li hog aphy. The ela i ely young combina ion o so li hog aphy wi h cas ing-, (mic oinjec ion) molding- and ho -embossing-app oaches o e mul iple p ocessing pa hs o he ab ica ion o mic o luidic de ices. 169-174 This makes a wide ange o (polyme ic) ma e ials accessible o mic o luidics and o e s g ea po en ial o he ab ica ion o mic o luidic de ices wi h ma e ial p ope ies ailo ed he speci ic expe imen al needs while main aining he g ea design lexibili y o apid p o o yping. As an example, apid-p o oyping-based PDMS-s amps and - empla es can be used o emboss he mic ochannel s uc u es in o a wide ange o he moplas ic ma e ials such as polys y ene (PS), poly(me hyl me hac yla e) (PMMA), cyclic ole in copolyme s (COC) and THV ( luo ina ed e polyme : e a luo oe hylene, hexa luo op opylene and inylidene luo ide). 148,150,175-182 Fu he , he use o ab ica ion echniques ha a e based on mic oinjec ion molding o ho embossing can g ea ly inc ease he p oduc ion speed o mic o luidic de ices o indus ial scales due o he ela i ely as embossing s ep in he empla e eplica ion p ocess. 163,169,170,172,183,184 Nex o Kap on wi h i s g ea p ope ies o X- ay applica ions, a numbe o X- ay compa ible window ma e ials ha e been es ed including polyme hylme hac yla e (PMMA), cyclic ole in copolyme s (COC), nega i e pho o esis SU-8 (Mic oChem) o polyp opylene (PP). I is possible o ab ica e hin (ca. 250 µm o less) and mic os uc u ed ilms ou o hese ma e ials o which he backg ound sca e ing is e y simila o ai . 9,163 F om hese 34 examples, he moplas s like PMMA o COC can easily be s uc u ed by so - o ho embossing while SU-8 can be mic os uc u ed di ec ly by UV-li hog aphy. 9,184,185 Mul iple uses o COC as a ma e ial o mic o luidic de ices ha e been demons a ed. 184 COC-based mic o luidic de ices also o e a g ea po en ial when i comes o he in eg a ion o unc ional elemen s in o he de ice. These in eg a ed ea u es o COC de ices a e mani old and include Au-elec odes o elec ochemis y, elemen s o elec och oma og aphy, o PDMS-based con ol laye s o he sc eening o lyo opic phases. 186-188 Fu he , he su ace chemis y o COC-based de ices can be modi ied by UV-g a ing p ocedu es and he e o e op imized o he pa icula expe imen al needs. 189 Among he X- ay- ela ed examples is a compu e -coupled high h oughpu sc eening se up by A le h e al. ha uses mic o luidic de ices o he as SAXS analysis o p o ein- olding unde he in luence o changing bu e condi ions. 176,179 Fu he , he au oma iza ion po en ial o mic o luidic de ices is demons a ed by using X- ay-CDs ha a e combined wi h au oma ed sample posi ioning a synch o on beamlines (PETRA III, DESY, Hambu g). These ‘SAXS-LabDiscs’ a e o a ional mic o luidic de ices ha a e based on cen i ugal low p inciples o he combina o ic mixing and sc eening o samples. 125,180,190-194 I has also been shown ha COC-based mic o luidic de ices a e sui able o mic o luidic g azing-incidence small-angle X- ay sca e ing (GISAXS) expe imen s. 35 Howe e , a downside i his polyme ic ma e ial is he limi ed sol en compa ibili y and, based on i s he moplas ic na u e, he lack o esis ance agains highe empe a u es. 163,177,180,184,189,195 As desc ibed abo e, he ad en o new echniques and ma e ials o e s new pa hs o he ab ica ion o mic o luidic de ices ha a e ailo ed o he expe imen al equi emen s. A di e en de ice ab ica ion app oach will be discussed in he nex sec ion wi h he gene a ion o liquid je s ha a oid any ma e ials in he X- ay beam pa h. 2.4 Mic o luidic liquid je de ices o nex gene a ion X- ay sou ces Today’s de elopmen s in synch o on echnology con inually push he peak b illiance o he X- ays, enabling as measu emen s as well as ime- esol ed in si u expe imen s. 7 Consequen ly, adia ion-induced sample deg ada ion and he maximum X- ay dose s a o become he limi ing ac o s o expe imen s a hese acili ies. As desc ibed abo e, one way a oid he dose p oblem is o scan di e en spa ial posi ion wi h mic o ocused beams, o use (mic o luidic) con inuous low sys ems. Howe e , each o hese al e na i es is limi ed o ce ain sample ypes. 45,142 Ano he pa h o o e coming he X- ay dose p oblems becomes a ailable wi h he ad en o highly in ense and ul asho -pulsed X- ay ee elec on lase s (XFEL). 6,49 These ee- 35 elec on lase sou ces cu en ly gene a e X- ay pulses a a es up o 120 Hz (LCLS, SLAC, S an o d, USA) while u u e acili ies will e en gene a e up o ca. 27,000 pulses pe second ha a e bundled in 10Hz bunch ains (Eu opean XFEL, Hambu g, Ge many, in 2015/2016). These X- ay pulses a e so in ense ha a ull di ac ion pa e n o nanome e - sized c ys als is collec ed wi hin a single sho and ha he X- ay beams “a e capable o des oying any hing in hei pa h”. 6,50,196 Du ing he illumina ion wi h a single ligh pulse, he sample explodes and u ns in o a glowing plasma (ca. 60,000 K). 196. Due o he ul asho em osecond pulse leng h howe e , he X- ay pulses ou un he explosion p ocess, ‘ eezing’ he a om posi ions in space. 49,50 An analogue example om he mac oscopic wo ld o his ‘ eezing’ p inciple is he highspeed lash pho og aphy wo k done by Ha old Edge on (MIT, Camb idge, USA). Fo example, his pho og aph “Cu ing he Ca d Quickly” (1964, 1 µs exposu e) shows how a bulle -sepa a ed play ca d le i a es in ai a i s o iginal posi ion. This concep o ‘di ac ion be o e des uc ion’ is o g ea impo ance o he s udy o ul a as p ocesses and he cha ac e iza ion o a wide ange o X- ay sensi i e samples. 49 As an example, he s uc u e de e mina ion o ansmemb ane p o eins is impo an o medicine and he li e sciences because his in o ma ion can help o unde s and he signalling o cells o mechanisms o diseases. 6 Howe e , he s uc u e de e mina ion o p o eins is a e y challenging ask because hese and many o he biological samples a e only a ailable in small amoun s and/o end no o o m single c ys als o su icien dimensions o adi ional X- ay (mic o-)c ys allog aphy. 6 I is much easie o g ow nanome e -sized p o ein c ys als, bu hese canno be analyzed a adionional X- ay sou ces due o hei small size and he discussed dose limi . The ela ion be ween he a e age in ensi y o a di ac ion peak and he c ys al olume is app oxima ely p opo ional. 45 Hence, he equi ed beam in ensi y o a 20 µm c ys al is 1000- old highe han o a 200 µm c ys al. 45 Theo e ical and expe imen al esul s show ha he c i ical dose o a success ul s uc u al e alua ion is mainly dic a ed by he c ys al size, despi e c yogenic a emp s o measu ing p o ein c ys als ha a e cooled below 100 K. 45 Recen ly, em osecond X- ay p o ein nanoc ys allog aphy has been demons a ed o he high- esolu ion cha ac e iza ion o pho osys em I&II and he model p o ein lysozyme, showing ha he X- ay dose limi can be o e come as desc ibed abo e. 6,197-199 As a consequence o he eno mous X- ay pulse in ensi y, s a ically moun ed samples o expe imen al en i onmen s o lowing samples ha a e based on closed geome ies a e incompa ible wi h hese 4 h gene a ion X- ay sou ces. The e o e, special sample en i onmen s a e equi ed ha deli e he samples in mid-ai and unde acuum condi ions while being as sample e icien as possible. 53,200 These kinds o sample en i onmen s o XFELs and how mic o luidics o e s g ea po en ial o he adequa e deli e y o samples will be discussed in his chap e . Cu en ly, con inuously eplenishing sample s eams a e gene a ed using ae odynamic lens pa icle injec o s o liquid je s using a glass-based capilla y-in-capilla y design. 53,200 Since he 36 glass capilla ies a e widely used and sui able o liquid samples, his chap e will ocus only on hese. The p inciple o he gene a ion o liquid je is based on a gas shea h which shapes a liquid s eam and has i s been demons a ed in a pla e-o i ice geome y. 201,202 La e , his concep has been ans e ed o glass capilla ies which un essen ially clogging ee due o hei gas-dynamic i ual nozzle design (GDVN). 53 In o he wo ds, he p essu ed gas o ms a liquid s eam and a oids any wall con ac o he liquid. 53 This esul s in a e y s able and eliable sys em o he gene a ion o nano- o mic ome e -sized liquid je s ha equi e only small amoun s o sample (down o ca. 100 µl h-1). 203,204 The main d awback o his glass capilla y design is he complex ab ica ion p ocess ha in ol es s eps like lame polishing o he ip, i s g inding as well as alignmen o he inne and ou e capilla y. 53,203-205 These s eps equi e he manual skills and a en ion o a lab wo ke which in insically esul s in geome ic a ia ions o he nozzles. I is he e o e e y ha d o impossible o exac ly ep oduce a a ge ed design o e en au oma e he p ocess. This is a majo issue because he gene a ed luid dynamics o he liquid je and, hence, minimum low a es and liquid je diame e s s ongly depend on he geome ic and expe imen al pa ame e s. 205-207 The ab ica ion p ocedu e is one key poin whe e mic o luidic de ices shine, due o hei as and easy ab ica ion and highly ep oducible design which is based on es ablished so - li hog aphical echniques. 55-57 In his hesis, mic o luidic chip-based de ices a e p esen ed ha p oduce liquid je s wi h µm-diame e s (20 down o 2 µm, o e en 940 nm) a e y low low a es (down o 150 µl h-1) unde a mosphe ic o acuum condi ions. These mic o luidic liquid je de ices a e also based on he gas dynamic i ual nozzle (GDVN) design which enables eliable and essen ially clogging- ee je ing o e long pe iods o ime. 53 The lexibili y in mic ochannel design con ol is demons a ed by he easy in eg a ion o addi ional mic o luidic ea u es, such as je -in-je low ocusing, which could enable new in si u expe imen s a XFELs, o dense a ays o mul iple adjacen liquid je nozzles on a single de ice, wi hou he need o addi ional p oduc ion s eps. Hence, he po en ial o simpli ying and up-scaling he ab ica ion o mic o-nozzles o he gene a ion o liquid je s is demons a ed. The mic o luidic liquid je sys em is highly ele an o he es ablishmen o mic o luidics a XFELs because hese de ices deli e he sample con inuously, eliably and e icien ly in a moshpe ic o unde acuum condi ions, as illus a ed in Fig.6. 136 37 Figu e 6 Combining he liquid je p inciple wi h mic o luidics. (A) Illus a ion o he building block p inciple o unc ional mic o luidic ools ha can be combined and s acked using he mic o luidic liquid je de ice p inciple. (B) The X- ay beam hi s he liquid je in his illus a ion o he expe imen al se up. Mic o luidic de ices a e ad an ageous because each liquid je de ice con ains a dense a ay mul iple mic o luidic GDNV-nozzles which enable as nozzle changes and, consequen ly, educe (expensi e) down imes a he X- ay ee elec on lase s. (Image adap ed and ex ended om 6) 38 2.5 Conclusions In conclusion, he combina ion o mic o luidics wi h mic o locused X- ays is a aluable expe imen al me hodology o he s udy o as in si u expe imen s. While his ield is s ill eme ging, a wide ange o de ice ypes is al eady a ailable. The on-going de elopmen o mic o ab ac ion echniques and ad en o ma e ials o he p oduc ion o X- ay compa ible mic o luidic de ices add o he g ea po en ial and his echnique’s u u e applica ions. Addi ionally, he a ie y o o he X- ay imaging and spec oscopic echniques could ex end he expe imen al oppo uni ies o mic o luidics a X- ay sou ces e en u he . Fu u e de elopmen s o mic o luidic sys ems and highly b illian X- ay sou ces, such as synch o ons o XFELs, could soon lead o he undamen al unde s anding o nuclea ion and g ow h p ocesses o in eg a ion o he high- h oughpu sc eening o p o eins ha yields ull h ee- dimensional as well as dynamics in o ma ion abou hese species o e en whole cells; wi h impo an insigh s o he na u al and li e sciences. 39 40 3 Theo e ical Fundamen als 3.1 Sel -o ganiza ion o amphiphiles One o na u e’s undamen al p inciples is sel -assembly. 208 While echnical sys ems a e ypically o ganized by men, he sel -o ganiza ion o na u al sys ems is based on in e nal p ocesses on he e y small scale. This bo om-up app oach begins wi h single unc ional molecules ha build up and o ganize hemsel es o la ge s uc u al elemen s. The sel - o ganiza ion o a sys em induces p ope ies ha i did no ha e be o e, like i.e. o de , and his p ocess can be desc ibed by he heo y o spon aneous symme y b eaking. 208,209 Fo example, his mechanism can be obse ed in na u e wi h he o ma ion o lipid double memb anes o cells o he sel -assembly o micelles and liquid c ys als. 209-211 Fo molecules o be able o o m such assemblies o highe o de , hei s uc u e and unc ional elemen s need o mee ce ain molecula p e equisi es. 208 These molecules need o possess (and uni e) bo h long ange epulsi e and sho ange a ac i e o ces in o de o be capable o o ming s uc u ed domains, as illus a ed in Fig.7. Figu e 7 Illus a ion o amphiphile building blocks wi h long ange epulsi e and sho ange a ac i e o ces and hei sel -assembly. ( om 208, Copy igh WILEY-VCH) Long ange epulsi e o ces can appea as Coulomb epulsion, chemical incompa ibili y o hyd ophobic in e ac ions while co alen bonds o he local conse a ion o elec oneu ali y a e examples o sho ange a ac i e o ces. 208 These p e equisi es a e me by molecules such as lipids o amphiphilic block copolyme s o which he la e can o m a wide a ie y o supe s uc u es as illus a ed in Fig8. 1,208 Con inuous phases a e o med in bulk o in lyo opic phases, and colloids like sphe ical o cylind ical micelles o esicles a e ypically o med by mic ophase sepa a ion in dilu e solu ions. 1,208 41 3.3 Solu ion o non-linea p oblems Fo linea sys ems, he inpu is p opo ional o he ou pu . Mos sys ems which a e o in e es o science and hei applica ions a e non-linea due o he complexi y o na u e. Thei success ul simula ion is o g ea in e es because p ecious esou ces such as ma e ial cos s, man-powe and de elopmen ime can be sa ed. Fu he , hese sys ems can lead o in e es ing non-linea scien i ic p oblems o which a ew examples will be discussed below. 3.3.1 Fluid s uc u e in e ac ion As men ioned ea lie , PDMS is a so polyme ic ma e ial ha de o ms elas ically unde o ce, like i.e. high p essu es. Depending on he applica ion, he mechanical p ope ies o his elas ome can also be bene icial o a down side. On he one hand, he de o ma ion unde high p essu es and high low a es can esul in ounded channels and educed he wall con ac o a hyd odynamically ocused liquid s eam. 137 On he o he hand, his p essu e-induced de o ma ion (a >2-3 ba ) can al e he mic ochannel geome y which could make he p edic ion o low condi ion mo e complica ed. Wi h he help o compu e -based CFD simula ions (COMSOL Mul iphysics 4.2a) i is possible o sol e hese p oblems nume ically based on he ini e elemen me hod. This app oach couples he mic ochannel geome y de o ma ion wi h he luid low ield which a e inc emen ally a ec ing each o he in a non- linea way. This ou ine esul s in he dynamically s able s a e o his p oblem. The heo e ical backg ound as well as he sol ed models, which a e good ag eemen wi h he eal expe imen al demons a ions, a e desc ibed in one o he publica ions o his hesis (see chap e 7.5). 137 3.3.2 Shea o ces and Non-New onian luids Ano he non-linea p oblem occu s when one has o p edic he low o non-New onian luids. These luids change iscosi y unde he in luence o o ces like shea o elonga ion. Gene ally speaking, he iscosi y o non-New onian luids can inc ease (shea hickening, dila an ) o dec ease (shea hinning, pseudoplas ic). The wo mlike polyme ic micelle solu ions which ha e been used in his hesis a e examples o shea hinning non-New onian luids. Thei change o iscosi y unde shea can be desc ibed by he Cole-Cole- (o C oss- ) equa ion which is gi en by: 247-249 wi h he ze o shea iscosi y , he high-shea iscosi y , he in e nal elaxa ion ime and he powe law exponen cha ac e izing he shea hinning be ween and .13,135 The luid low o non-New onian luids can now be calcula ed by coupling his equa ion wi h he abo e-desc ibed Na ie -S okes equa ions o an incomp essible luid and unning he FEM-based CFD-simula ion. As desc ibed in he publica ions o his hesis (see chap e 7.1 & 7.2), he simula ed low and p edic ed shea and ex ensional o ces a e also in good ag eemen wi h he expe imen al esul s (SAXS, µPIV, pola iza ion mic oscopy) and can be 48 applied i.e. o he p edic ion o pe pendicula pa icle o ien a ion in con ined geome ies. 13,135 3.3.3 Two-phase sys ems and liquid je s Ano he example o a non-linea sys em, which has been s udied in his hesis, is he luid low o wo-phase sys ems (see chap e 7.3). 136 The gas-dynamic i ual nozzle p inciple is a wo-phase low sys em which uses a p essu ed-gas shea h o he gene a ion o liquid je s. 53 This design p e en s wall con ac o he liquid and nozzles o his ype un essen ially clogging- ee while consuming only small amoun s o sample. 53,54 The non-linea i y o his sys em lies in he coupled low ields o each luid. When he liquid en e s he nozzle geome y, whe e he gas is al eady lowing, he liquid’s su ace shape is a ec ed by he gas low. A he same ime, i s p esence al e s he gas low dynamics, e c. The heo e ical backg ound and he simula ion a e desc ibed in a pape o his hesis. 136 In his pape , a ime- esol ed model enables o begin he simula ion wi h easily de inable s a ing condi ions. The CFD-simula ion o his coupled in e ac ion inc emen ally leads o a s able equilib ium s a e and a s able gas-shaped liquid je . 54,207 The esul ing simula ed liquid je shape and diame e a e ound o be in good ag eemen wi h he expe imen al esul s om high speed ideo mic oscopy. 136 This CFD-model allows he e y de ailed analysis o he whole sys em, including je shape, p essu e g adien s, shea a es, eloci ies a e e y simula ed posi ion o each luid. I one only needs o es ima e he liquid je diame e , he e is also an al e na i e, analy ical app oach which has been desc ibed o a pla e-o i ice con igu a ion, as illus a ed in Fig.11. 201 Assuming cylind ical coo dina es o he axis o he liquid je , he cusplike meniscus a he inle is pulled owa ds he nozzle o i ice by he p essu e g adien ha is gene a ed by he gas s eam. 201 This p essu e di e ence and angen ial iscous s ess o he gas shea h leads o he o ma ion o a hin liquid h ead wi h he adius as illus a ed in Fig.11. 201,206 The a e aged momen um equa ion o his case is gi en by: 201 wi h wi h he low a e , he liquid p essu e and he su ace ension s ess . The liquid e apo a ion will be neglec ed as well as he iscous ex ensional e m which is negligible compa ed o he kine ic ene gy e m. This holds ue o many low a es o s able liquid h eads. 201 Thus, he abo e a e aged momen um equa ion can be simpli ied o: 49 Figu e 11 Illus a ion o a gene al pla e-o i ic nozzle design o he gene a ion o liquid je s and he in ol ed geome ic pa ame e s. The liquid exi s he capilla y wi h he diame e a a low a e o . I is shaped by he p essu ed gas shea h ( , le & igh a ows) along he pa h owa ds he nozzle and as i passes he nozzle wi h he diame e . The diame e and he s abili y o he esul ing liquid je depend on he gi en geome ic pa ame e s as well as o he luid dynamic p ope ies such as luid densi y , iscosi y o su ace ension . (Image om 206, Copy igh Ame ican Ins i u e o Physics). Assuming high p essu e g adien s owa ds he nozzle and ha he con ining nozzle ape u e is o he o de o hinne han i s diame e , his equa ion can be in eg a ed. 201 This in eg a ion yields a simple and uni e sal exp ession o he je diame e ha is gi en by This o mula is independen o he geome ical pa ame e s (like i.e. inle - and ou le diame e s, inle - o-oule -dis ance, e c.), liquid-gas su ace ension and liquid and gas iscosi ies. 201 The alidi y o his exp ession has been demons a ed expe imen ally o pla e- o i ice con igu a ions and glass capilla y se ups. 201,206 Ou expe imen al esul s show ha he e is also a good ag eemen o his o mula wi h he liquid je diame e s ha a e gene a ed in mic o luidic gas-dynamic i ual nozzles (see chap e 7.3). 136 When i comes o he lowes possible low a e howe e , he nozzle shape has a decisi e e ec . 206 I has been demons a ed ha he minimum low a e o s able je ing can be con olled by adjus ing he geome y. 206 Fo h ee di e en nozzle diame e s ( ), he ollowing g aph (Fig.12) shows he minimum low a es ( ) which a e plo ed agains he dis ance ( ) be ween he liquid inle and he nozzle ape u e. This g aph clea ly shows ha his - a io is e y sensi i e. Consequen ly, a p ecise and ep oducible nozzle design con ol is essen ial o he op imiza ion o liquid je sys em and hei in eg a ion as a sample en i onmen a high in ensi y X- ay sou ces. 6 50 Figu e 12 The minimum low a e o s able je ing o wa e a di e en combina ions o nozzle dis ance and nozzle diame e s wi h 100 µm (ci cles), 200 µm (squa es) and 400 µm ( iangles). The p essu e di e ence was se o 250 mba . (Image om 206, Copy igh Ame ican Ins i u e o Physics). The egimes o s able liquid je s (s eady je ing) can be mapped ou using Reynolds- and Webe numbe diag ams as illus a ed in Fig.13. 206 The b eakup ansi ion lines depend on he speci ic nozzle geome ies & luid p ope ies and a e he e o e only alid o a speci ic case, as indica ed in Fig.13 bo om. The ca e ul mapping o a scena io’s je -b eakup pa ame e combina ions allows o iden i y i s b eakup ype. In case o a pla e-o i ice con igu a ion, he ollowing equa ions can be used o he pa ame e con e sion: 206 and wi h wi h he densi y o he luid , he low a e , he je adius , he iscosi y , he su ace ension and he p essu e di e ence . These equa ions can also be combined in a adius-independen o m i he exac alue o he p essu e di e ence is known: and As desc ibed abo e, he Reynolds numbe desc ibes he a io be ween ine ial and iscous o ces while he Webe numbe is he a io be ween he ine ia o he luid compa ed o i s su ace ension. 10,213 A s able je is obse ed i he luid ou uns he ins abili ies con ec i ely. 201,206,207,250,251 I he gas low is oo high o a gi en liquid s eam o example, he je u ns in o a sp ay which co esponds o a local&global s abili y- o local&global ins abili y- ansi ion. This ‘ igh - o-le ’ ansi ion is ma ked as a blue in e sec ed line in Fig.13. 201,206,207,250,251 51 I he gas p essu e a a gi en low a e is lowe ed slowly (‘ op- o-bo om’), he b eakup co esponds o a local s abili y- o local ins abili y- ansi ion while being globally s able. In his case a con inuous d ople ain a a cons an equency is obse ed wi h a s eady liquid column in he ange o he nozzle opening. This b eakup ype is ma ked by he ed do ed line in Fig.13 and is also known as he Leib-Golds ein limi . 250,251 Fu he lowe ing o he p essu e inally leads he global ins abili y (Fig.13 bo om). 201,206,207,250,251 Unde ce ain condi ions when he nozzle geome ies become e y small, as in mic o luidic liquid je de ices and e y small je s, he ela i e in luences o he luid’s su ace ension and o he shea om as - lowing gas s eams on he liquid su ace inc ease. Hence, he unde lying assump ions (see momen um equa ion discussion abo e) a e no neglec able anymo e and he Re- and We-numbe con e sions can become inaccu a e. 201,206 I is he e o e p e e able o ely on quali a i e esul s o he iden i ical ion o b eakup ypes. This can be achie ed by eco ding he je b eakup ansi ion using highspeed came as. This has been demons a ed in he a ached pape on mic o luidic liquid je sys ems, i.e. o he column- leng h o liquid je s o d ople ains in he local ins abili y egime (see chap e 7.3). 136 Figu e 13 Reynolds- and Webe numbe diag ams and je b eakup ypes a minimum low a es. The liquid je s a e obse ed in he s eady je ing egime. The ed do ed line desc ibes he Leib-Golds ein ansi ion (local s abili y o local ins abili y). 250,251 The blue in e sec ed line ma ks he global s abili y o ins abili y ansi ion. The op diag am ma ks he di e en je b eakup ansi ion egions. Sp aying can ypically be obse ed in he global ins abili y egime while a con inuous d ople s eams can be an example o he local ins abili y b eakup ype. The bo om g aph illus a es how he cu es in he Re-We-space shi wi h changing nozzle geome ies. (Images om 206, Copy igh Ame ican Ins i u e o Physics). 52 3.4 Small angle X- ay sca e ing 3.4.1 P e ace Small angle X- ay sca e ing (SAXS) is one o oday’s mos impo an expe imen al echniques o he cha ac e iza ion o so condensed ma e and colloidal sys ems. This chap e will gi e an o e iew o e i s heo e ical backg ound and desc ibe he undamen al X- ay sca e ing p inciples. A mo e de ailed desc ip ion can be ound in he o iginal, ci ed li e a u e on which his chap e is based. 12,252-263 3.4.2 In oduc ion The sca e ing o elec omagne ic wa es is a e y help ul ool o he cha ac e iza ion o colloids and polyme s. The undamen al se up o a sca e ing expe imen is illus a ed in Fig. 14. Fig. 14 Illus a ion o he undamen al elemen s o a sca e ing expe imen . The X- ay sou ce (X) emi s ligh which passes a collima ion sys em (C). The X- ays hen hi he sample (S) and he sca e ing pa e n is eco ded using a de ec o (D) while he p ima y beam is abso bed by he beams op (B). (Image adap ed om 258) The X- ays, which should be as monoch oma ic as possible, a e emi ed om a sou ce (X), like a synch o on o o a ing anode. The X- ays pass a collima ion sys em (C) be o e hey hi he sample (S). This in e ac ion causes he sample’s elec ons o esona e and he induced dipoles emi seconda y wa es o he same equency. This sca e ing p ocess is conside ed o be elas ic because he inciden pho ons ha e he same ene gy as he sca e ed ones. The sca e ed X- ays a e cohe en and he phases o hese seconda y wa es di e om each o he due o he di e en spa ial posi ions o he sca e ing elec ons. Consequen ly, in e e ence o hese seconda y wa es occu s and he esul ing sca e ing pa e ns a e hen eco ded using 2D digi al de ec o s (i.e. Pila us, FReLoN, Ma CCD) o al e na i e eco ding echniques such as 1D coun ing de ices o image pla es. The in e e ence o he sca e ed X- ays and he e o e he sca e ing pa e n is cha ac e is ic o he gi en sample’s s uc u e. The undamen al p inciple o he wa es’ in e e ence can be desc ibed by he B agg equa ion. 253 A sca e ing e lex, which is an in ensi y maximum, can be ound whe e he wa es’ pa h di e ence is an in ege mul iple o he wa eleng h which leads o cons uc i e 53 in e e ence. This p inciple is illus a ed in Fig. 15. which clea ly shows ha he pa h di e ence is de e mined by he sca e ing angle and he dis ance o he sca e ing planes . Fig. 15 Illus a ion o he B agg equa ion wi h he inciden X- ays and , he la ice dis ance and he hal sca e ing angle . The pa h di e ence o he wa e is gi en by . (Image adap ed om 263) The esul ing sca e ing pa e n o a sample is an angle-dependen in ensi y dis ibu ion ha is cha ac e is ic o he sample’s s uc u e. Gi en by he ange o he (small) sca e ing angles o his me hod, he size ange o s uc u es ha a e s udied by SAXS ypically lies be ween 1 nm and 100 nm. This is also why his expe imen al echnique has become e y popula o he cha ac e iza ion o sample sys ems such as polyme s, colloids, so condensed ma e and nanoma e ials. 3.4.3 The sca e ing ec o Fo a be e unde s anding o he in e e ence phenomena one can imagine he in e ac ion be ween he X- ay beam and wo sca e ing elec ons a he posi ions and . The ec o which desc ibes he dis ance be ween hem is gi en by as illus a ed in Fig. 16. Fig. 16 Phase ela ion be ween he wo sca e ing cen e s and and he geome ic cons uc ion o he sca e ing ec o . (Image adap ed om 258) All in e e ences o wa es ha o igina e om sca e ing e en s sum up wi h espec o hei ampli ude and phase. Due o he equali y o elec ons when i comes o X- ay sca e ing, only 54 he loca ion-dependen phase di e ence needs o be conside ed which is gi en by This equa ion con ains he wa e ec o s which a e gi en by o he inciden wa e, while he sca e ed wa e is desc ibed by wi h he uni ec o s ( ) in -di ec ion. The absolu es o hese ec o s a e gi en by Conside ing he angle o he X- ays , he wa e ec o s can be combined o cons uc he sca e ing ec o which is gi en by The geome ic ep esen a ion o his ela ion is shown in Fig. 16. The phase shi can be desc ibed by he mul iplica ion o his ec o wi h . The sca e ing cu e is ecei ed by plo ing he measu ed in ensi y agains he sca e ing ec o ’s absolu e alue which is gi en by The absolu e alue o can also be exp essed as which is based on he ollowing equa ion whe e he ela ion be ween , and has a simila i y o he B agg equa ion. 3.4.4 The sca e ing pa e n The cohe en sca e ed X- ay wa es in e e e wi h each o he and he ampli ude o he esul ing wa e is gi en by wi h he sca e ing leng h . Fo SAXS he sca e ing leng h o an elec on is gi en by255 55 wi h he elemen a y cha ge , he elec on’s mass and he speed o ligh . Consequen ly, he sca e ing leng h o an a om wi h he o de ing numbe is gi en by The cumula i e sca e ing o a sample is he sum o all sca e ed wa es. The e o e, he collec i e sca e ing ampli ude is ecei ed by in eg a ion and he sca e ing leng h o a single pai o sca e ing cen e s is eplaced by he densi y dis ibu ion o all sca e ing cen e s . This is desc ibed by he ollowing o mula. Since his equa ion’s ma hema ical o m is a Fou ie ans o ma ion, he sca e ing ampli ude and he sca e ing cen e densi y dis ibu ion a e a pai o Fou ie ans o ms. Fu he , his equa ion links he eal space (wi h he ec o ) o he ecip ocal space (wi h he sca e ing ec o ). Howe e , he sca e ing ampli ude is expe imen ally no accessible because only he sca e ing in ensi y is measu ed. Thei ela ion is de ined as In o he wo ds, he sca e ing in ensi y is he ime-a e aged squa e o he absolu e alue o he sca e ing ampli ude . I is ime-a e aged, as indica ed by he poin ed b acke s , because he measu emen is long compa ed o he sys em’s dynamics. Ano he ele an ma hema ical ope a ion is he con olu ion o wo unc ions and i is gi en by I is commonly known ha a con olu ion o and in he eal space co esponds o he mul iplica ion o hei Fou ie ans o ms and in he ecip ocal space: 263 The applica ion o his con olu ion heo em on he complex sca e ing ampli ude and he sca e ing in ensi y yields As b ie ly desc ibed abo e, he sca e ing ampli ude and he sca e ing cen e densi y a e a pai o Fou ie ans o ms. The abo e con olu ion heo em also shows ha his is also ue o he sca e ing in ensi y and he pai co ela ion unc ion . The sca e ing in ensi y esul s om he squa e o he absolu e alue o he sca e ing ampli ude while he pai dis ibu ion unc ion esul s om he sel -con olu ion o he sca e ing cen e densi y . 56 In he o m o a Fou ie ans o m can also be w i en as264 which is he spa ially a e aged in ensi y o a s a is ically iso opic sys em wi hou any long ange o de , such as dilu e pa icles. He e, is he au o-co ela ion unc ion which is gi en by and he ollowing equa ion highligh s he ela ion be ween he pai dis ibu ion unc ion and he au o-co ela ion The ollowing Fig. 17 illus a es hese ela ions be ween eal and ecip ocal space g aphically. Fig. 17 G aphical ep esen a ion o he ma hema ical ope a ions ha link he sca e ing ampli ude , he sca e ing cen e densi y , he sca e ing in ensi y and he pai co ela ion unc ion . (Image adap ed om 258) The Fou ie - ans o ma ion is ully e e sible in bo h di ec ions while he squa e o he absolu e alue and he sel -con olu ion a e no . The wan ed sca e ing cen e densi y is no ex ac able om he expe imen ally measu ed sca e ing in ensi y because he phase in o ma ion is missing; his is also known as he phase p oblem. This is why wo di e en app oaches ha e been de eloped o ecei e he sca e ing cen e densi y . The so-called indi ec me hod is based on he modeling o he sca e ing cen e densi y using spline unc ions. 265-270 The splines a e ans o med o he measu emen space and i ed o he sca e ing cu e. The desi ed sca e ing cen e densi y is inally ecei ed by he decon olu ion o hese i ed spline unc ions. 271,272 Ano he app oach is he model-based o di ec me hod. This app oach uses a gi en s uc u e wi h a known sca e ing cen e densi y ha is Fou ie - ans o med o ecei e he 57 wi h This p o ile shape smoo hly ansi ions om a Lo en z peak shape a e y small - alues o a Gauß shape a la ge alues o The peak wid h is de e mined by he domain size which is gi en by The co ela ion unc ion is gi en by which also desc ibes he de ia ion, known as he Debye-Walle ac o , om he ideal la ice posi ion based on he ollowing equa ion In his o mula is he mean squa e de ia ion and is he nea es neighbo dis ance. 3.4.10 Pa icle o ien a ion dis ibu ion In he abo e discussion, he o ien a ion o aniso opic pa icles, such as cylind ical micelles o wo mlike micelles, has been assumed o be iso opic. This esul s in ing-like, iso opic sca e ing pa e ns on wo-dimensional de ec o s which can simply be ep esen ed as adially a e aged sca e ing cu es. I he pa icles in a sample a e o ien ed howe e , he esul ing sca e ing pa e ns a e also aniso opic and bo h componen s o he sca e ing ec o ( and ) need o be conside ed in he analysis. A good expe imen al example o aniso opic sca e ing pa e ns a e small angle X- ay sca e ing s udies o shea -o ien ed wo mlike micelles in small mic o luidic channel geome ies, as illus a ed in Fig. 22. 64 Fig. 22 (A) The sample is lowing hough he mic ochannel geome y and passes a na ow sec ion ( ed box). (B) 3D illus a ion o he X- ay mic obeam ha passes he mic ochannel o lowing wo mlike micelles (hexagonal closes packing, de e mined by SAXS) and he esul ing SAXS-pa e n which is cap u ed using a digi al de ec o (Pila us 300K, Dec is). The pa allel o pe pendicula o ien a ion o he wo mlike micelles is con olled by a ying he expe imen al pa ame e s, such as he channel geome y, he low a e o he sample concen a ion.(Adap ed om 13, Copy igh PNAS) The wo mlike micelles in he mic ochannel a e o ien ed acco ding o he low-induced shea and ex ensional o ces which a e con olled by he expe imen al pa ame e s such as mic ochannel geome y, low speed and pa icle concen a ion. Due o he highly ep oducible low condi ions, his se up enables he co ela ion be ween s uc u al- and o ien a ion in o ma ion om SAXS s udies wi h luid dynamic s udies om o he me hods like high speed ideo analysis, pa icle image elocime y o pola iza ion mic oscopy. Ins ead o spli ing he sca e ing ec o in o i s componen s ( and ) i is mo e bene icial o use he pola coo dina e sys em. 252 He e, he sca e ing ec o is desc ibed by i s absolu e alue and he angle be ween cylinde axis and he sca e ing ec o . This esul s in he ollowing equa ion o he sca e ing in ensi y o o ien ed cylind ical micelles258 wi h he o m ac o , he s uc u e ac o , he ac ion o micelles wi h he angle and he numbe o micelles . The o m ac o is u he de ined by 65 The dis ibu ion unc ion desc ibes he cylind ical micelle o ien a ion wi h he angle be ween he cylinde axis wi h he base ec o and he di ec o which de ines he di ec ion o he shea ield. The exp essions and a e yielded om he ac o iza ion o he sca e ing ampli ude in o i s c oss sec ion- and leng h-con ibu ion. The poin ed b acke s indica e he a e aging ac oss he co esponding size dis ibu ions o cylinde s and adii. As desc ibed abo e, i is possible o exp ess o ypical bock copolyme s, ha ha e a co e-shells-s uc u e and a densi y-p o ile o , by using hype geome ic unc ions. is gi en by wi h and he a ios o he adii ( ) and densi ies ( ) compa ed o he shell. These a e gi en by The o m ac o calcula ion equi es a de ini ion o he ela ion be ween he angles and , as shown in Fig. 23. Fig. 23 Th ee-dimensional illus a ion o he ec o sphe e. (Image om 252,258, Copy igh Else ie ). 66 In his igu e he di ec o is de e mined by he angles and . The base ec o s o he cylinde axes a e posi ioned on a cone ha is di ec ed owa ds . The e o e, he ange o angles o is gi en by and he in eg a ion o he unc ion leads o The ec o is calcula ed by using he o a ion ma ix based on he ollowing equa ion The ec o s and a e gi en by and The o a ion ma ix is de ined by wi h Las ly, is gi en by The o ien a ion o aniso opic pa icles can be desc ibed by a ange o dis ibu ion unc ions ha a e gi en by 67 wi h he pa ame e which can ake on alues be ween ze o and in ini y. The g aphs o hese di e en dis ibu ion unc ions a e shown in Fig. 24. Fig. 24 The g aphs o di e en dis ibu ion unc ions. (Image om 252, Copy igh Else ie ). Among hese unc ions, he Onsage and Maie -Saupe dis ibu ions can be poin ed ou because hey a e e y impo an o he desc ip ion o pa icle o ien a ions in lyo opic and he mo opic liquid-c ys alline sys ems. The dis ibu ion unc ions a e no malized using he ollowing ac o The esul ing mean de ia ion angle be ween cylinde s and he di ec o , which can ake on alues be ween 0° and 90°, is gi en by A e y gene al unc ion is he Lague e dis ibu ion which changes om a Gauß unc ion (a 68 ) o a Hea iside unc ion (a e y la ge ). The o de pa ame e can ake on alues be ween 0 and 1. Fo a known dis ibu ion unc ion i is gi en by The abo e equa ions a e alid o dilu ed sys ems and wi h aising concen a ions i becomes necessa y o ake he s uc u e ac o in o accoun as desc ibed by an-de -Schoo 284 wi h he cylinde concen a ion and wi h 69 70 4 Me hods and Techniques 4.1 Pho oli hog aphy Ea ly ma e ials ha ha e been used o he ab ica ion o mic o luidic de ices include glass o silicon. 78 These ma e ials we e used because a wide ange o mic os uc u ing echniques we e eadily a ailable om he ield o semiconduc o echnology and because hese ma e ials o e esis ance agains a wide ange o sol en s. While glass is addi ionally op ically anspa en , silicon is opaque which limi s mic oscopic applica ions o he la e ma e ial. A downside o bo h ma e ials is he need o expensi e clean oom en i onmen s du ing hei p ocessing and he use o agg essi e chemicals which makes i an expensi e and esou ce in ensi e p ocess. 57,285 The as and e ec i e ab ica ion o mic o luidic de ices in a sho ime became possible h ough he p og ess in he a eas o pho o- and so li hog aphy. 56 Polyme -based ma e ials o e mo e applica ion iendly p ope ies which is why hey ha e gained impo ance du ing he pas yea s. 230,286 One example o hese a e pho o esis s like SU-8 which a e used in pho oli hog aphy. 287 This echnology enables he c ea ion o mic os uc u es based on speci ic echnical d awings which a e c ea ed using compu e aided design (CAD) so wa e. By using he compu e -designed pho o masks, pho o esis s a e selec i ely exposed o he gene a ion o mic os uc u es. 56 These mic os uc u es a e hen eplica ed using so ma e ials such as polyme s o c ea e mic os uc u ed s amps o mic o luidic de ices which is why his echnology is called so li hog aphy. 55 Due o he p ecise con ol o e design ea u es and he sho ened sys em op imiza ion eedback loop, his p ocess enables (simula ion-based) apid p o oyping. 57 Fu he mo e, he eplica ion empla es, also called mas e s, can be e-used mul iple imes wi hou quali y loss enabling mass p oduc ion and low ab ica ion cos s. 288 The mic os uc u ing o he pho o esis happens by selec i e exposu e using pho o masks. 57 In case o a nega i e pho o esis like SU-8, he non-c oss-linked a eas o he pho o esis emain soluble and will be emo ed in he de elopmen s age o he li hog aphic p ocess. The de elopmen ba h ypically con ains 80 o 100% 2-me hoxy-1-me hyle hyl ace a e solu ion and is, simila o he pho o esis s, comme cially a ailable (m -DEV 600, Mic ochem). Only he insoluble c oss-linked a eas o he pho o esis emain on he subs a e, ypically a polished la silicon wa e , and will be used as a mas e empla e o he subsequen eplica ion s eps. 55,56 The esolu ion o so li hog aphy is limi ed by di ac ion and, hence, dependen on he wa e leng h o he ligh sou ce. 56 The ange o a ailable pho o esis ma e ials is only small due o he speci ic p ocessing equi emen s. 71 Figu e 25 Illus a ion o he apid p o o yping concep which desc ibes he sho design- eedback-loop ha enables he as op imiza ion o mic o luidic sys ems. (Image adap ed and modi ied om 56) Figu e 26 Fab ica ion o SU-8 mas e empla es. A polished silicon wa e is spin coa ed using he nega i e pho o esis SU-8. Using selec i e UV-exposu e wi h an emulsion ilm mask, he mic ochannel geome y (i.e. mixing c oss o hyd odynamic ocusing) is ecei ed a e he de elopmen p ocess which emo es he non-c oss-linked ma e ial. (Image om 13, Copy igh PNAS) The pho o esis EPON SU-8 is a ma e ial ha is widely used o he ab ica ion o mic o luidic de ices. De eloped by Shell Chemicals, his highly- unc ionalized monome (69% in -bu y olac one) is capable o in e molecula c oss-linking which enables he o ma ion o h ee-dimensional s uc u es wi h nanome e esolu ion. 289-292 Figu e 27 S uc u al o mula o EPON SU-8. 222,293,294 72 The mechanism o his c oss-linking is a ca ionic ing opening polyme iza ion (ROMP) o he epoxide ha is s a ed by pho o-ini ia ed a yl sul onium sal s. The li hog aphic p ocess in ol es a UV-exposu e s ep ha leads o he o ma ion o luo oan imonic acid ( ). This is a s ong Lewis acid ha is o med in he p esence o ia ylsul onium hexa luo oan imona e (3.3% solu ion in p opylene ca bona e, 4-Me hyl-1,3- dioxolan-2-one) and p o on dono s like he o ganic sol en . 295-297 This eac ion mechanism is illus a ed in Fig.28. Figu e 28 Pho ochemical pa hways du ing he gene a ion o luo oan imonic acid ( ).297,298 Since he pho o ini ia o is consumed in an al e na i e eac ion pa h du ing he UV- exposu e, i is c i ically impo an o adjus he ligh dose acco dingly. O he wise, he equi ed ini ia o is no a ailable in su icien amoun s o he acid-induced epoxide c oss- linking du ing he subsequen “pos exposu e bake” s ep, which is he hea ing o he exposed pho o esis . The eac ion mechanism o he ca ionic ing opening polyme iza ion o he SU-8 pho o esis is shown in Fig.29. Figu e 29 Ca aly ic eac ion mechanism du ing he c oss-linking o SU-8. 299 The epoxide g oup is p o ona ed and he nucleophilic a ack o he hyd oxyl g oup o ano he SU-8 molecule leads o he opening o he epoxide ing. The ca alys is egene a ed by he elimina ion o a p o on and he cycle epea s i sel which leads o a c oss-linked ne wo k. 73 A key bene i o he mic os uc u ing echnique called so li hog aphy is i s p ecise con ol o he mic ochannel geome ies. I is he e o e possible o combine di e en expe imen al echniques in a complemen a y way. Consequen ly, he exac ep oduc ion o mic ochannel geome ies is c ucial o he combined use o di e en expe imen al echniques. Acco dingly, mic o ocus SAXS-s udies, mic oscopic expe imen s, such as pola iza ion mic oscopy o mic opa icle image elocime y (µPIV), and compu a ional luid dynamics (CFD) simula ions ha e been used in an in eg a ed app oach ha has been applied in he success ul explana ion o he pe pendicula e-o ien a ion e ec (see chap e 7.1). We could show by pola iza ion mic oscopy, µPIV-expe imen s and addi ional CFD simula ions ha he pe pendicula o ien a ion o wo mlike micelles is he esul o he in e play be ween he x- o ien ing shea - and y-o ien ing ex ensional o ces; wi h x along he low di ec ion and y along he widening o he ape ing. The e-o ien a ion o aniso opic pa icles has been in es iga ed in u he de ail in he ollowing pape (see chap e 7.2). The es ablished sys em o jus -desc ibed complemen a y analysis me hods has been applied o a sys ema ic pa ame e sc eening, including he ape ing a ios (-wid h, -leng h), he low speeds and concen a ions. Fu he , he heological pa ame e s ha e been a ied sys ema ically using CFD-simula ion ha a e based on he sample’s expe imen ally measu ed heological pa ame e s. The sys em’s esponse o hese pa ame e a ia ions has been analyzed and quan i ied ca e ully. A ela i e anking o he desc ibed sys em-con olling pa ame e s could be de i ed om he combined expe imen al and heo e ical esul s. The esul s o his wo k (chap e s 7.1 & 7.2), and he pe pendicula e-o ien a ion e ec in gene al, a e o g ea impo ance o applica ion ha equi e o ien a ion con ol, such as injec ion molding, ibe spinning o p ocessing o composi e ma e ials. The second (chap e 7.2) pape also desc ibes an imp o ed e sion o he closed-channel X- ay compa ible mic o luidic de ice ha is based on a ecen ly published pape . 132 These de ices a e made o he UV-cu able adhesi e NOA81 (see chap e 2.1 and 4.2) as he de ice ma e ial. Howe e , he published design also has wo weak poin s. The i s is ela ed he ubing-in e ace which glues PDMS on o he la NOA81-de ice, punches a hole h ough he comple e s uc u e and seals i again he sel -adhesi e Tape. This educes he p essu e esis ance as well as he sol en compa ibili y o he de ice (see chap e 2.1). The second law is he lack o he de ice’s heigh con ol. Since a co ec backg ound sub ac ion is essen ial o SAXS-expe imen s, especially o dilu e o weakly-sca e ing samples, a homogenous de ice heigh is o g ea impo ance, as is he o e all ma e ial hickness o he mic o luidic de ice. The imp o ed ab ica ion o NOA81-de ices is desc ibed in he expe imen al sec ion while i s illus a ion is p esen ed in he supplemen al sec ion o his pape (chap e 7.2). The i s imp o emen in ol es he in eg a ion o b idging s uc u es ha a e used o con olling he de ice heigh by adjus ing he pho o esis laye heigh h ough spin-coa ing. This leads o e y e en and ex emely hin and s able mic o luidic chips, educing he X- ay backg ound 80 signal and imp o ing he signal ansmission. The second imp o emen is ela ed o he ubing in e ace. Du ing he molding o eplica ion s ep, a ubing is punched h ough he (PDMS- )mic os uc u e and se es as a empo a y empla e, enabling he di ec connec ion o he ubing o he inished eplica. As a consequence, he sol en compa ibili y o NOA81 is ully main ained and many pola and unpola sol en s, which a e incompa ible wi h PDMS, can now be used in his pu ely NOA81-based de ices wi hou issues. Addi ionally he p essu e esis ance o hese de ices is imp o ed and can e en be enhanced by glueing ubing o he de ice. Fu he , hese imp o ed de ices un mo e eliably, ha e a smalle de ice oo p in enabling dense a ays o mic ochannels, be e op ical p ope ies and lowe backg ound signal compa ed o PDMS-Kap on-de ices. Abo e all, he ab ica ion is easie due o he p ecise heigh con ol and much quicke due o he as UV-exposu e imes (unde one minu e) compa ed o ens o minu es, o e en hou s, o he mally cu ed PDMS. Ano he mic o luidic de ice ha is also made o his ma e ial (NOA81) is desc ibed in chap e 7.4. This a ia ion is closed and no ac i ely lowing, bu he e o e only e y small sample amoun s a e equi ed o he SAXS-measu emen o shea ed samples. Wi h only 2- 5 µl o luid olume, he nanopa icle lyo opic gel sample is applied o a millime e -sized pa ch o he mic os uc u ed g id (14 µm spacing) while being shea ed in he p ocess. The mic og id is hen sealed agains e apo a ion using Kap on ape. Hence, his de ice demons a es he high sample e iciency and ease o use when i comes o expe imen s a synch o ons. Addi ionally, his wo k was essen ial o he imp o emen and op imiza ion o so li hog aphic ab ica ions echniques ela ed o NOA81 ha lead o he success ul ab ica ion o high- esolu ion NOA81-based X- ay compa ible mic o low de ices which a e p esen ed in chap e 7.2. The nex pape desc ibes a mic o luidic liquid je sys em (see chap e 7.3). As an example o an open mic o luidic sample en i onmen , i is designed o expe imen s a highly b illian X- ay sou ces o X- ay ee elec on lase s. The liquid je is gene a ed based on he gas-dynamic i ual nozzle design (GDVN) which ep esen s he cu en s a e o he a a XFELs. 6 Fo his eason his mic o luidic de ice uns essen ially clogging- ee and highly eliable o e long pe iods o ime. 53 The goal o his pape was o c ea e a mic o luidic liquid je design ha is easy and as o ab ica e, since he cu en glass capilla y ab ica ion is a complex and manually challenging p ocedu e. The c ea ion o glass-based nozzles equi es a skilled p oduce because he ab ica ion in ol es manual s eps like g inding, lame-polishing and alignmen o he capilla ies. In con as , he he e p esen ed mic o luidic liquid je nozzles can easily be eplica ed by using s anda d PDMS-de ice ab ica ion s eps which a e e y as o lea n. The mic o luidic de ices u he o e he bene i o a pa allelized nozzle design which enables o c ea e complex je -in-je - ocusing geome ies o comple e a ays o mul iple nozzle simul aneously in one ab ica ion sequence. Consequen ly he de ice oo p in is e y small which enables as nozzle changes by simply swi ching o he adjacen nozzles. 81 The ab ica ion o hese mic o luidic liquid je de ices became possible due o ou ad ances in mul i-laye ed so li hog aphy. This enables he ab ica ion o 3D-mic ochannels by sealing wo ma ching hal es o mic os uc u ed PDMS ha a e ea ed wi h an ai plasma. Since he mic ome e -p ecise alignmen would be manually challenging, he eplica ion empla es al eady inco po a e alignmen s uc u es by design. These snap-in alignmen s uc u es could be in eg a ed h ough he mul i-laye ed mic os uc u e design and hey wo k simila ly o Lego® building blocks, locking he mic os uc u es o he wo PDMS-hal es in hei igh posi ion du ing he sealing s ep. A key bene i o his mic o luidic de ice o e he s a e o he a glass capilla ies is he highly ep oducible design ha is enabled h ough so li hog aphy. This mic ome e -p ecise con ol o e he nozzle geome ies is c i ically impo an o he op imiza ion o he liquid je s, especially a small low a es. As desc ibed in chap e 2.4, he geome ic pa ame e s s ongly a ec he minimum low a e o s able je ing which is he key ac o o he high sample e iciency o hese de ices. Taking ad an age o his design con ol, he ele an pa ame e s, such as liquid low a e and p essu e di e ence, we e a ied o s udy he liquid je dynamics. We ound ha he mic o luidic liquid je diame e s can adjus ed wi h g ea con ol by a ying he p essu es & low a es and ha hey a e in good ag eemen wi h a ailable analy ical exp essions o he p edic ion o je diame e s in pla e-o i ice geome ies. 201 This has also been e i ied by in-si u en i onmen al scanning elec on mic oscopy o he liquid je exi ing he nozzle. Fu he mo e, he a ia ions o he abo e pa ame e s allowed o con ol he je b eakup ype. These je b eakup ansi ions ha e been s udied using highspeed ideo mic oscopy and we also ound ha he s able je ’s column leng h can be con olled o e a wide leng h scale. The je shape could be p edic ed success ully by using ime- esol ed non-linea 3D CFD- simula ions ha desc ibe he wo-phase low o luids. The simula ed shape o his heo e ical liquid je is in good ag eemen wi h he expe imen al mic oscopic esul s. The je ’s 3D shape was also s udied by using con ocal lase scanning mic oscopy. Since his s udy in ol ed an enhanced design ha inco po a ed a je in je hyd odynamic ocusing geome y, be o e he liquid je is shaped by he gas shea h, he luid low wi hin he liquid je could also be s udied using in hese expe imen s. In e es ingly, we ind ha he low o he wo dye solu ions ( hodamine B and luo escein) is in e ed and he inne ocused liquid s eam can be obse ed a he ou side o he liquid je as i exi s he nozzle geome y. This in e sion migh o igina e om he in e sion o speeds a he ou e bounda ies o he liquid ha lead o he gene a ion o o ices coun e ing he liquid low. 207 The eloci y p o ile is pa abolic inside he mic ochannel, he e o e as es in he liquid s eam’s cen e and slowes close o he walls, his scena io changes quickly as he liquid exi s he mic ochannel. Suddenly, he as es liquid eloci y can be ound a he ou side whe e he as - lowing gas is in con ac wi h he liquid. This kind o di usion-con olled mixing inside he liquid je could 82 enable new kinds o expe imen s a XFELs, such as he em osecond-pulse di ac ion o in si u nuclea ion and g ow h p ocesses. The mic o luidic concep s and li hog aphic mic o ab ica ion echniques ha e also been used o c ea ing solu ions o applica ions ha a e no ela ed o X- ay expe imen s. As p esen ed in chap e 7.5, a CFD-model has been c ea ed which desc ibes non-linea in e ac ion be ween he luid and he mic ochannel walls unde high p essu es and high low a es. The CFD-simula ions showed how PDMS-mic ochannels wi h high aspec a ios de o med unde hese unusual condi ions owa ds a ounded shape. This ound shape was in good ag eemen wi h he mic oscopic images. We u he ound ha his ound shape o mic ochannel educes he wall con ac o lowing species which we e mixed by hyd odynamic ocusing be o e en e ing his ound passage. The e o e, clogging o he de ice can be s ongly educed o elimina ed. The c ea ion o his CFD-model ha e lead o deepe insigh s abou he simula ion so wa e and i s capabili ies and laid a ounda ion o he o he non-linea CFD-models desc ibed abo e. So li hog aphy is essen ial o he wo k o he p esen ed pape s o which he ounda ion has been laid wi h he ollowing wo pape s. The pape desc ibed in chap e 7.6 desc ibes he combina ion o nano echnology and mic os uc u ing echniques o de elopmen s owa ds senso ic applica ions. This wo k equi ed he c ea ion o deep, high aspec a io, SU-8 mic ochannels. The op imiza ion he pa ame e s du ing he UV-li hog aphic p ocess lead o be e -de ined and imp o ed mic ochannel wall geome ies o all mic o luidic de ices o his hesis. The wo k desc ibed in chap e 7.7 ocuses on inc easing he so li hog aphic esolu ion o mic os uc u ed empla es and an op imized eplica ion he eo . This op imiza ion esul ed in e y small and well-de ined le e s o high esolu ion ha se ed as a mic o con ac p in ing s amp o he de ined deposi ion o sphe ical polyelec oly e b ushes. This inc eased li hog aphic esolu ion lead o he capabili y o ab ica ing e y small mic o luidic ea u es, such as na ow ape ings (chap e 7.1 & 7.2), na ow g ids (chap e 7.4) and small nozzles (chap e 7.3). Hence, he li hog aphic de elopmen s o hese wo pape s played a c ucial ole o he success ul applica ion o so li hog aphy in he o he pape s by inc easing he quali y o high aspec a io mic ochannel walls as well as he inc easing esolu ion o e y small mic os uc u es. The ollowing chap e will desc ibe he au ho s’ con ibu ion o hese pape s, ollowed by he pape s in he subsequen chap e . 83 Re e ences 1. M. An onie i and S. Fö s e , Ad anced Ma e ials, 2003, 15, 1323–1333. 2. S. Fö s e , Topics in Cu en Chemis y, 2003, 226, 1–28. 3. S. Fö s e , B. Be on, H. P. Hen ze, E. K äme , M. An onie i, and P. Lindne , Mac omolecules, 2001, 34, 4610–4623. 4. I. Robinson, G. G uebel, and S. Moch ie, New J. Phys., 2010, 12. 5. T. Na ayanan, Cu Opin Colloid in, 2009. 6. H. N. Chapman, P. F omme, A. Ba y, T. A. Whi e, R. A. Ki ian, A. Aquila, M. S. Hun e , J. Schulz, D. P. DePon e, U. Weie s all, R. B. Doak, F. R. N. C. Maia, A. V. Ma in, I. Schlich ing, L. Lomb, N. Coppola, R. L. Shoeman, S. W. Epp, R. Ha mann, D. Rolles, A. Rudenko, L. Fouca , N. Kimmel, G. Weidenspoin ne , P. Holl, M. Liang, M. Ba helmess, C. Caleman, S. Bou e , M. J. Bogan, J. K zywinski, C. Bos ed , S. Baj , L. Gump ech , B. Rudek, B. E k, C. Schmid , A. Hömke, C. Reich, D. Pie schne , L. S üde , G. Hause , H. Go ke, J. Ull ich, S. He mann, G. Schalle , F. Schoppe , H. Sol au, K.-U. Kühnel, M. Messe schmid , J. D. Bozek, S. P. Hau-Riege, M. F ank, C. Y. Hamp on, R. G. Sie a, D. S a odub, G. J. Williams, J. Hajdu, N. Timneanu, M. M. Seibe , J. And easson, A. Rocke , O. Jönsson, M. S enda, S. S e n, K. Nass, R. And i schke, C.-D. Sch ö e , F. K asniqi, M. Bo , K. E. Schmid , X. Wang, I. G o johann, J. M. Hol on, T. R. M. Ba ends, R. Neu ze, S. Ma chesini, R. F omme, S. Scho b, D. Rupp, M. Adolph, T. Go kho e , I. Ande sson, H. Hi semann, G. Po de in, H. G aa sma, B. Nilsson, and J. C. H. Spence, Na u e, 2011, 470, 73–77. 7. D. H. Bilde back and P. Elleaume, Jou nal o Physics B: …, 2005. 8. H. A. S one, A. D. S oock, and A. Ajda i, Annu. Re . Fluid Mech., 2004, 36, 381–411. 9. S. Kös e and T. P ohl, Mod. Phys. Le . B, 2012, 26, 1230018. 10. T. Squi es and S. Quake, Re . Mod. Phys., 2005, 77, 977–1026. 11. J. Als-Nielsen and D. McMo ow, Elemen s o Mode n X- ay Physics, Wiley, Chiches e , 2011. 12. O. Gla e and O. K a ky, Academic P ess, 1982. 13. M. T ebbin, D. S einhause , J. Pe lich, A. Bu e , S. V. Ro h, W. Zimme mann, J. Thiele, and S. Fö s e , PNAS, 2013, 110, 6706–6711. 14. S. M. Tahe i, S. Fische , M. T ebbin, S. Wi h, J. H. Sch öde , J. Pe lich, S. V. Ro h, and S. Fö s e , So Ma e , 2012, 8, 12124. 15. L. Daube sies, J. Leng, and J.-B. Salmon, Lab Chip, 2013, 13, 910. 16. C. Riekel, P. Engs öm, and C. Ma in, J. o Mac omolecula Sc., Pa B, 1998, B37, 587– 599. 17. R. J. Da ies, M. Bu ghamme , and C. Riekel, Synch o on Radia ion in Na u al Science, 2006, 5, 96–99. 18. S. V. Ro h, R. Döh mann, M. Dommach, M. Kuhlmann, I. K öge , R. Geh ke, H. Wal e , C. Sch oe , B. Lengele , and P. Mülle -Buschbaum, Re . Sci. Ins um., 2006, 77, 085106. 84 19. A. Timmann, R. Döh mann, T. Schube , H. Schul e-Sch epping, U. Hahn, M. Kuhlmann, R. Geh ke, S. V. Ro h, A. Sch opp, C. Sch oe , and B. Lengele , Re . Sci. Ins um., 2009, 80, 046103. 20. C. Riekel, M. Bu ghamme , and R. Da ies, IOP Con . Se .: Ma e . Sci. Eng., 2010, 14, 012013. 21. A. Bu e , A. Ro hki ch, R. Döh mann, V. Kö s gens, M. M. Abul Kashem, J. Pe lich, G. He zog, M. Schwa zkop , R. Geh ke, P. Mülle -Buschbaum, and S. V. Ro h, J. Synch o on Rad (2012). 19, 647-653 [doi:10.1107/S0909049512016895], 2012, 1–7. 22. A. Ma el, M. Bu ghamme , R. Da ies, E. DiCola, P. Panine, J.-B. Salmon, and C. Riekel, Biomic o luidics, 2008, 2, 024104. 23. A. O en, S. Kös e , B. S u h, A. Snigi e , and T. P ohl, J Synch o on Rad, 2005, 12, 745–750. 24. A. Snigi e , V. Kohn, I. Snigi e a, and B. Lengele , Na u e, 1996. 25. B. S u h, AIP, 2004, ol. 705, pp. 804–807. 26. C. G. Sch oe and B. Lengele , Phys. Re . Le ., 2005, 94, 054802. 27. M. Mülle , C. Riekel, R. Vuong, and H. Chanzy, Polyme , 2000, 41, 2627–2632. 28. C. Riekel and R. J. Da ies, Cu Opin Colloid in, 2005, 9, 396–403. 29. R. J. Da ies, M. Bu ghamme , and C. Riekel, Mac omolecules, 2007, 40, 5038–5046. 30. A. Ma el, M. Bu ghamme , R. J. Da ies, E. Di Cola, C. Vend ely, and C. Riekel, J. Am. Chem. Soc., 2008, 130, 17070–17074. 31. C. Riekel, M. Rossle, D. Sapede, and F. Voll a h, Na u wissenscha en, 2004, 91, 30–33. 32. C. Riekel, B. Madsen, D. Knigh , and F. Voll a h, Biomac omolecules, 2000, 1, 622–626. 33. C. Riekel, M. Mülle , and F. Voll a h, Mac omolecules, 1999, 32, 4464–4466. 34. M. E. Kinahan, E. Filippidi, S. Kös e , X. Hu, H. M. E ans, T. P ohl, D. L. Kaplan, and J. Wong, Biomac omolecules, 2011, 12, 1504–1511. 35. J. F. Moulin, S. V. Ro h, and P. Mülle -Buschbaum, Re . Sci. Ins um., 2008, 79, 015109. 36. N. S ibeck, A. A. Cama illo, U. Nöchel, C. Sch oe , M. Kuhlmann, S. V. Ro h, R. Geh ke, and R. K. Baye , Mac omolecula Chemis y and Physics, 2006, 207, 1139–1149. 37. N. S ibeck, U. Nöchel, and A. Almenda ez-Cama illo, Mac omolecula Chemis y and Physics, 2008, 209, 1976–1982. 38. J. M. Feldkamp, M. Kuhlmann, S. V. Ro h, A. Timmann, R. Geh ke, I. Shakh e do a, P. Pau le , S. K. Fila o , R. S. Bubno a, and C. G. Sch oe , Phys. S a us Solidi (a), 2009, 206, 1723–1726. 39. C. G. Sch oe , M. Kuhlmann, S. V. Ro h, R. Geh ke, N. S ibeck, A. Almenda ez- Cama illo, and B. Lengele , Appl. Phys. Le ., 2006, 88, 164102–164102–3. 40. M. Kuhlmann, J. M. Feldkamp, S. V. Ro h, and C. G. Sch oe , 2008, pp. 570–573. 41. J. M. Feldkamp, Disse a ion, 2009, 1–136. 42. A. C. Kak and M. Slaney, IEEE P ess, 1988. 43. A. Gonzalez and C. Na e, Ac a C ys allog D Biol C ys allog , 1994, 50, 874–877. 44. R. Hende son, P oceedings o he Royal Socie y B: Biological Sciences, 1990, 241, 6–8. 85 45. S. Cusack, H. Bel hali, A. B am, M. Bu ghamme , A. Pe akis, and C. Riekel, Na . S uc . Biol., 1998, 5, 634–637. 46. L. Pollack, M. Ta e, N. Da n on, J. Knigh , S. G une , W. Ea on, and R. Aus in, PNAS, 1999, 96, 10115–10117. 47. L. Pollack, M. W. Ta e, A. C. Finne ock, C. Kalidas, S. T o e , N. C. Da n on, L. Lu io, R. H. Aus in, C. A. Ba , S. M. G une , and S. Moch ie, Phys. Re . Le ., 2001, 86, 4962–4965. 48. R. Russell, I. S. Mille , M. W. Ta e, L. W. Kwok, B. Naka ani, S. M. G une , S. Moch ie, V. Pande, S. Doniach, D. He schlag, and L. Pollack, P oceedings o he Na ional Academy o Sciences, 2002, 99, 4266–4271. 49. R. Neu ze, R. Wou s, D. an de Spoel, E. Wecke , and J. Hajdu, na u e, 2000, 406, 752–757. 50. L. Lomb, T. Ba ends, S. Kassemeye , A. Aquila, S. Epp, B. E k, L. Fouca , R. Ha mann, B. Rudek, D. Rolles, A. Rudenko, R. Shoeman, J. And easson, S. Baj , M. Ba helmess, A. Ba y, M. Bogan, C. Bos ed , J. Bozek, C. Caleman, R. Co ee, N. Coppola, D. DePon e, R. B. Doak, T. Ekebe g, H. Fleckens ein, P. F omme, M. Gebha d , H. G aa sma, L. Gump ech , C. Hamp on, A. Ha mann, G. Hause , H. Hi semann, P. Holl, J. Hol on, M. Hun e , W. Kabsch, N. Kimmel, R. Ki ian, M. Liang, F. R. N. Maia, A. Meinha , S. Ma chesini, A. Ma in, K. Nass, C. Reich, J. Schulz, M. M. Seibe , R. Sie a, H. Sol au, J. C. Spence, J. S einb ene , F. S ella o, S. S e n, N. Timneanu, X. Wang, G. Weidenspoin ne , U. Weie s all, T. Whi e, C. Wunde e , H. Chapman, J. Ull ich, L. S üde , and I. Schlich ing, Phys. Re . B, 2011, 84, 214111. 51. A. Ba y, C. Caleman, A. Aquila, N. Timneanu, L. Lomb, T. A. Whi e, J. And easson, D. A nlund, S. Baj , T. R. M. Ba ends, M. Ba helmess, M. J. Bogan, C. Bos ed , J. D. Bozek, R. Co ee, N. Coppola, J. Da idsson, D. P. DePon e, R. B. Doak, T. Ekebe g, V. Else , S. W. Epp, B. E k, H. Fleckens ein, L. Fouca , P. F omme, H. G aa sma, L. Gump ech , J. Hajdu, C. Y. Hamp on, R. Ha mann, A. Ha mann, G. Hause , H. Hi semann, P. Holl, M. S. Hun e , L. Johansson, S. Kassemeye , N. Kimmel, R. A. Ki ian, M. Liang, F. R. N. C. Maia, E. Malme be g, S. Ma chesini, A. V. Ma in, K. Nass, R. Neu ze, C. Reich, D. Rolles, B. Rudek, A. Rudenko, H. Sco , I. Schlich ing, J. Schulz, M. M. Seibe , R. L. Shoeman, R. G. Sie a, H. Sol au, J. C. H. Spence, F. S ella o, S. S e n, L. S üde , J. Ull ich, X. Wang, G. Weidenspoin ne , U. Weie s all, C. B. Wunde e , and H. N. Chapman, Na u e Pho on, 2011, 6, 35–40. 52. J. Ull ich, A. Rudenko, and R. Moshamme , Annu Re Phys Chem, 2012, 63, 635–660. 53. D. P. DePon e, U. Weie s all, K. Schmid , J. Wa ne , D. S a odub, J. C. H. Spence, and R. B. Doak, J. Phys. D: Appl. Phys., 2008, 41, 195505. 54. A. M. Gañán-Cal o, D. P. DePon e, M. A. He ada, J. C. H. Spence, U. Weie s all, and R. B. Doak, Small, 2010, 6, 822–824. 55. Y. Xia and G. M. Whi esides, Annual e iew o ma e ials science, 1998, 28, 153–184. 56. G. M. Whi esides and Y. Xia, Angew. Chem. In . Ed., 1998, 37, 550–575. 86 57. D. C. Du y, J. C. McDonald, O. J. A. Schuelle , and G. M. Whi esides, Anal. Chem., 1998, 70, 4974–4984. 58. G. M. Whi esides, Na u e, 2006, 442, 368–373. 59. C.-M. Ho and Y.-C. Tai, Annu. Re . Fluid Mech., 1998, 30, 579–612. 60. A. G oisman, M. Enzelbe ge , and S. Quake, Science, 2003, 300, 955–958. 61. J. Thiele, Uni e si y o Bay eu h, 2011. 62. G. M. Whi esides and A. D. S oock, Physics Today, 2001, 54, 42. 63. A. S oock and G. M. Whi esides, Elec opho esis, 2002, 23, 3461–3473. 64. A. Manz, N. GRABER, and H. M. WIDMER, Senso s and Ac ua o s B: Chemical, 1990, 1, 244–248. 65. A. R. Aba e, M. B. Romanowsky, J. J. Ag es i, and D. A. Wei z, Appl. Phys. Le ., 2009, 94, 023503. 66. A. R. Aba e, T. Hung, P. Ma y, J. J. Ag es i, and D. A. Wei z, P oceedings o he Na ional Academy o Sciences, 2010, 107, 19163–19166. 67. T. Tho sen, S. J. Mae kl, and S. R. Quake, Science, 2002, 298, 580–584. 68. Y.-C. Tan, V. C is ini, and A. P. Lee, Senso s and Ac ua o s B: Chemical, 2006, 114, 350– 356. 69. W. Xu and S. J. Mulle , Lab Chip, 2011, 11, 435–442. 70. T. P ohl, Analyse, Manipula ion und Agg ega ion on biologischen Mak omolekülen im mik o luidischen Sche luss, De G uy e , 2010. 71. T. F anke and A. Wix o h, Phys. Unse e Zei , 2007, 38, 88–94. 72. A. R. Aba e, J. J. Ag es i, and D. A. Wei z, Appl. Phys. Le ., 2010, 96, 203509. 73. D. The iaul , S. R. Whi e, and J. A. Lewis, Na . Ma e ., 2003, 2, 265–271. 74. S. K. Sia and G. M. Whi esides, Elec opho esis, 2003, 24, 3563–3576. 75. C. Xi, D. L. Ma ks, D. S. Pa ikh, L. Raskin, and S. A. Boppa , P oceedings o he Na ional Academy o Sciences, 2004, 101. 76. D. S. Kim, S. H. Lee, T. H. Kwon, and C. H. Ahn, Lab Chip, 2005, 5, 739. 77. S. Ha d , K. S. D ese, V. Hessel, and F. Schön eld, Mic o luid Nano luid, 2005, 1, 108– 118. 78. S. Quake and A. Sche e , Science, 2000, 290, 1536–1540. 79. J. W. Hong and S. R. Quake, Na u e bio echnology, 2003, 21, 1179–1183. 80. J. W. Hong, V. S ude , G. Hang, W. F. Ande son, and S. R. Quake, Na u e bio echnology, 2004, 22, 435–439. 81. C. Lee, G. Sui, A. Eliza o , C. Shu, and Y. Shin, Science, 2005, 310, 1793–1796. 82. J.-C. Ba e , O. J. Mille , V. Taly, M. Ryckelynck, A. El-Ha ak, L. F enz, C. Rick, M. L. Samuels, J. B. Hu chison, J. J. Ag es i, D. R. Link, D. A. Wei z, and A. D. G i i hs, Lab Chip, 2009, 9, 1850. 83. H. Sha iee, M. B. Sano, E. A. Henslee, J. L. Caldwell, and R. V. Da alos, Lab Chip, 2010, 10, 438. 84. T. Q. Chas ek, K. Iida, E. J. Amis, M. J. Fasolka, and K. L. Bee s, Lab Chip, 2008, 8, 950. 87 85. K. Iida, T. Q. Chas ek, K. L. Bee s, K. A. Ca icchi, J. Chun, and M. J. Fasolka, Lab Chip, 2009, 9, 339. 86. C. Hany, H. Leb un, C. P ade e, J. Tou ain, and J.-C. Ba sale, Chemical Enginee ing Jou nal, 2010, 160, 814–822. 87. H. Pennemann, P. Wa s, S. J. Haswell, V. Hessel, and H. Lowe, O g. P ocess Res. De ., 2004, 8, 422–439. 88. S. Sei e and D. A. Wei z, Polyme , 2010, 51, 5883–5889. 89. E. Tuma kin and E. Kumache a, Chem. Soc. Re ., 2009, 38, 2161. 90. C. P ies , A. Quinn, A. Pos ma, A. N. Zelikin, J. Rals on, and F. Ca uso, Lab Chip, 2008, 8, 2182. 91. Y. Mo imo o, W.-H. Tan, Y. Tsuda, and S. Takeuchi, Lab Chip, 2009, 9, 2217–2223. 92. Y. Song, J. Ho mes, and C. S. S. R. Kuma , Small, 2008, 4, 698–711. 93. Y. Song, H. Mod ow, L. L. Hen y, C. K. Saw, E. E. Doomes, V. Palshin, J. Ho mes, and C. S. S. R. Kuma , Chem. Ma e ., 2006, 18, 2817–2827. 94. A. Jahn, J. Reine , W. V eeland, and D. DeVoe, Jou nal o Nanopa icle Resea ch, 2008. 95. P. Kallio and J. Kunco a, TEKES Technology e iew, 2004. 96. P. S. Di ich and A. Manz, Na Re D ug Disco , 2006, 5, 210–218. 97. W. Li, J. G eene , D. Voicu, and E. Kumache a, Lab Chip, 2009, 9, 2715. 98. R. S. Ramsey and J. M. Ramsey, Anal. Chem., 1997, 69, 1174–1178. 99. J. Pihl, M. Ka lsson, and D. T. Chiu, D ug Disco . Today, 2005, 10, 1377–1383. 100. A. R. Wheele , W. R. Th ondse , and R. J. Whelan, Analy ical …, 2003. 101. A. A. We dich, E. A. Lima, B. I ano , I. Ges, M. E. Ande son, J. P. Wikswo, and F. J. Baudenbache , Lab Chip, 2004, 4, 357. 102. R. T. Tu geon and M. T. Bowse , Anal Bioanal Chem, 2009, 394, 187–198. 103. I. K. Dimo , L. Basabe-Desmon s, J. L. Ga cia-Co de o, B. M. Ross, A. J. Ricco, and L. P. Lee, Lab Chip, 2011, 11, 845–850. 104. J. Melin and S. R. Quake, Annual e iew o biophysics and biomolecula s uc u e, 2007, 36, 213–231. 105. M. A. Unge , H. P. Chou, T. Tho sen, A. Sche e , and S. R. Quake, Science, 2000, 288, 113–116. 106. Y. Huang, P. Cas a a o, C.-C. Lee, and S. R. Quake, Lab Chip, 2006, 7, 24. 107. C. Zhang, J. Xu, W. Ma, and W. Zheng, Bio echnol. Ad ., 2006, 24, 243–284. 108. E. A. O esen, J. W. Hong, S. R. Quake, and J. R. Leadbe e , Science, 2006, 314, 1464–1467. 109. I. P. G. Ma shall, P. C. Blainey, A. M. Spo mann, and S. R. Quake, Applied and En i onmen al Mic obiology, 2012, 78, 8555–8563. 110. S. J. Pamp, E. D. Ha ing on, S. R. Quake, D. A. Relman, and P. C. Blainey, Genome Res., 2012, 22, 1107–1119. 111. J. Wang, H. C. Fan, B. Beh , and S. R. Quake, Cell, 2012, 150, 402–412. 112. F. E. Dewey, S. Pan, M. T. Wheele , S. R. Quake, and E. A. Ashley, Ci cula ion, 2012, 125, 931–944. 113. D. Pushka e , N. F. Ne , and S. R. Quake, Na u e bio echnology, 2009, 27, 847–850. 88 114. H. C. Fan, Y. J. Blumen eld, Y. Y. El-Sayed, J. Chueh, and S. R. Quake, YMOB, 2009, 200, 543.e1–543.e7. 115. R. A. Whi e, P. C. Blainey, H. C. Fan, and S. R. Quake, BMC Genomics, 2009, 10, 116. 116. M. Meie , R. Si , W. Pan, and S. R. Quake, Anal. Chem., 2012, 84, 9572–9578. 117. M. Meie , R. V. Si , and S. R. Quake, P oceedings o he Na ional Academy o Sciences, 2013, 110, 477–482. 118. C. L. Hansen, E. Sko dalakes, J. M. Be ge , and S. R. Quake, P oceedings o he Na ional Academy o Sciences, 2002, 99, 16531. 119. C. Hansen and S. R. Quake, Cu en opinion in s uc u al biology, 2003, 13, 538–544. 120. B. Zheng, J. D. Tice, and R. F. Ismagilo , Ad anced Ma e ials, 2004, 16, 1365–1368. 121. C. L. Hansen, S. Classen, J. M. Be ge , and S. R. Quake, J. Am. Chem. Soc., 2006, 128, 3142–3143. 122. B. Zheng, J. Tice, L. Roach, and R. Ismagilo , Angewand e Chemie (In e na ional ed. in English), 2004, 43, 2508. 123. J. D. Ng, P. J. Cla k, R. C. S e ens, and P. Kuhn, Ac a C ys (2008). D64, 189-197 [doi:10.1107/S0907444907060064], 2008, 1–9. 124. J. Leng and J.-B. Salmon, Lab Chip, 2009, 9, 24. 125. S. Emamzadah, T. J. Pe y, V. De Almeida, T. Nishimu a, J. Joly, J.-L. Fe e , and T. D. Halazone is, Ac a C ys (2009). D65, 913-920 [doi:10.1107/S0907444909021489], 2009, 1– 8. 126. L. Li and R. F. Ismagilo , Annu. Re . Biophys., 2010, 39, 139–158. 127. S. L. P. A. S. P. S. K. N. A. P. J. A. K. S Guha, 2011, 1–3. 128. M. Maeki, S. Yoshizuka, H. Yamaguchi, M. Kawamo o, K. Yamashi a, H. Nakamu a, M. Miyazaki, and H. Maeda, Anal Sci, 2012, 28, 65. 129. J. B. Knigh , A. Vishwana h, J. P. B ody, and R. H. Aus in, Phys. Re . Le ., 1998, 80, 3863–3866. 130. T. P ohl, A. O en, S. Kös e , R. Doo z, B. S u h, and H. M. E ans, Biomac omolecules, 2007, 8, 2167–2172. 131. S. Kös e , H. M. E ans, J. Y. Wong, and T. P ohl, Biomac omolecules, 2008, 9, 199–207. 132. M. E. B ennich, J.-F. Nol ing, C. Dammann, B. Nöding, S. Bauch, H. He mann, T. P ohl, and S. Kös e , Lab Chip, 2011, 11, 708. 133. C. Reich, M. B. Hoch ein, B. K ause, and B. Nickel, Re . Sci. Ins um., 2005, 76, 095103. 134. T. P ohl, F. Mugele, R. Seemann, and S. He minghaus, ChemPhysChem, 2003, 4, 1291– 1298. 135. M. T ebbin, S. Wi h, A. Ma k, C. Hanske, A. Bu e , G. San o o, S. Yu, J. Pe lich, S. V. Ro h, and S. Fö s e , in p epa a ion, 2013. 136. M. T ebbin, K. K üge , D. DePon e, J. Schulz, S. V. Ro h, H. N. Chapman, and S. Fö s e , in p epa a ion, 2013. 137. J. Thiele, M. Windbe gs, A. R. Aba e, M. T ebbin, H. C. Shum, S. Fö s e , and D. A. Wei z, Lab Chip, 2011, 11, 2362. 89 288. M. E. Sandison and H. Mo gan, J. Mic omech. Mic oeng., 2005, 15, S139–S144. 289. K. Y. Lee, J. Vac. Sci. Technol. B, 1995, 13, 3012. 290. M. M. A. A. K. Mohamed, 2010, 1–26. 291. J. Simčič, P. Pelicon, Z. Rupnik, M. Mihelič, A. Razpe , D. Jenko, and M. Maček, Nuclea Ins umen s and Me hods in Physics Resea ch Sec ion B: Beam In e ac ions wi h Ma e ials and A oms, 2005, 241, 479–485. 292. C. G. Khan Malek, Mic oelec onics Jou nal, 2002, 33, 101–105. 293. J. Zhang, K. Tan, G. Hong, L. Yang, and H. Gong, J. Mic omech. Mic oeng., 2001, 11, 20–26. 294. J. Zhang, K. Tan, and H. Gong, Polym. Tes ., 2001, 20, 693–701. 295. Y. Wang, M. Bachman, C. Sims, G. Li, and N. Allb i on, Langmui , 2006, 22, 2719– 2725. 296. J. Dek a and N. Hacke , J. Am. Chem. Soc., 1990, 112, 6004–6015. 297. Mic oChem NANOTM SU-8 O ganic Resin Solu ion, MSDS, 12.12.2001, www.cise.columbia.edu/clean/msds/su-8 esis .pd (1.07.2013), 2009. 298. Y. Wang, J.-H. Pai, H.-H. Lai, C. E. Sims, M. Bachman, G. P. Li, and N. L. Allb i on, J. Mic omech. Mic oeng., 2007, 17, 1371–1380. 299. Gue in, L. The SU-8 homepage, h p://www.geoci ies.com/gue inlj/ (17.07.2008). 2009. 300. X. Yan and P. Gu, Compu e -Aided Design, 1996. 301. E. Sachs, S. Allen, H. Guo, and B. Banos, 1997. 302. E. Sachlos and J. T. Cze nuszka, Eu Cell Ma e , 2003. 303. C. M. Cheah, C. K. Chua, C. W. Lee, C. Feng, and K. To ong, In J Ad Manu Technol, 2004, 25, 308–320. 304. A. Khademhosseini, PNAS, 2006, 103, 2480–2487. 305. C. J. Campbell, S. K. Smouko , K. J. M. Bishop, E. Bake , and B. A. G zybowski, Ad anced Ma e ials, 2006, 18, 2004–2008. 306. P. J. Ki son, M. H. Rosnes, V. Sans, V. D agone, and L. C onin, Lab Chip, 2012, 12, 3267–3271. 307. K. B. Ande son, S. Y. Lockwood, R. S. Ma in, and D. M. Spence, Anal. Chem., 2013, 85, 5622–5626. 308. V. D agone, V. Sans, M. H. Rosnes, P. J. Ki son, and L. C onin, Beils ein J O g Chem, 2013, 9, 951–959. 309. Y. Liao, J. Song, E. Li, Y. Luo, Y. Shen, D. Chen, Y. Cheng, Z. Xu, K. Sugioka, and K. Mido ikawa, Lab Chip, 2012, 12, 746. 310. A. D. Campo and C. G eine , J. Mic omech. Mic oeng., 2007, 17, R81–R95. 311. S. Jeon, J. U. Pa k, R. Ci elli, S. Yang, C. E. Hei zman, P. V. B aun, P. J. A. Kenis, and J. A. Roge s, P oceedings o he Na ional Academy o Sciences, 2004, 101, 12428. 312. J. A. Lewis and G. M. G a son, Ma e ials Today, 2004, 7, 32–39. 313. Y. L. Zhang, Q. D. Chen, H. Xia, and H. B. Sun, Nano Today, 2010. 314. B. Zimme , Nanosc ibe GmbH, 2013, 1–14. 315. MSDS: Sylga d(R) 184 silicone elas ome ke (base), DOW Co ning, 1.5, 2008, 2008. 316. MSDS: Sylga d(R) 184 silicone elas ome ke (cu ing agen ), DOW Co ning, 1.8, 2007, 2007. 96 317. H. Finkelmann and G. Rehage, Mac omol. Rapid Commun., 1980, 1. 318. B. K ucke, H. Zaschke, S. Kos omin, and V. Shibae , Ac a Polym., 1985, 36. 319. 2007, 1–3. 320. E. Delama che, A. Be na d, H. Schmid, A. Bie sch, B. Michel, and H. Biebuyck, J. Am. Chem. Soc., 1998, 120, 500–508. 321. G. M. Whi esides, E. Os uni, S. Takayama, X. Jiang, and D. Ingbe , Annu. Re . Biomed. Eng., 2001, 3, 335–373. 322. N. B. C ame , J. P. Sco , and C. N. Bowman, Mac omolecules, 2002, 35, 5361–5365. 97 98 6 Indi idual con ibu ion o join publica ions The esul s which a e p esen ed in he scien i ic pape s o his hesis a e based on he collabo a i e wo k wi h o he s. The con ibu ion o e e y co-au ho will be speci ied in u he de ail below and he co esponding au ho will be indica ed wi h an as e isk (*). Chap e 7.1 This wo k is published in PNAS, 2013, 110, 6706–6711 and ea u ed as a “Technical Highligh ” in DESY Pho on Science 2011 and on he on pages o he DESY- and Uni Bay eu h web pages; en i led: "Aniso opic pa icles align pe pendicula o he low di ec ion in na ow mic ochannels". by Ma in T ebbin, Dagma S einhause , Jan Pe lich, Adeline Bu e , S ephan V. Ro h, Wal e Zimme mann, Julian Thiele*, and S ephan Fö s e *. (sha ed co esponding au ho ship) I pe o med mos o he expe imen s & da a analysis, c ea ed he non-linea non-New onian low CFD-simula ion model; and w o e he manusc ip oge he wi h Julian Thiele and S ephan Fö s e . Dagma S einhause and Julian Thiele helped wi h he expe imen s a he synch o on beamlines a which Jan Pe lich, Adeline Bu e and S ephan V. Ro h p o ided he echnical suppo . Wal e Zimme mann was in ol ed in scien i ic discussions and helped wi h he co ec ion o he manusc ip . Fu he , S ephan Fö s e supe ised he p ojec . Chap e 7.2 This wo k is p epa ed as a manusc ip ha is eady o submission; en i led: "Mic o luidic SAXS o he high- h oughpu sc eening and co ela ion o complex luid beha io wi h s uc u al in o ma ion". by Ma in T ebbin, Sebas ian Wi h, And es Ma k, Ch is oph Hanske, Adeline Bu e , Gonzalo San o o, Shun Yu, Jan Pe lich, S ephan V. Ro h, and S ephan Fö s e *. I pe o med mos o he expe imen s & da a analysis, c ea ed he non-linea non-New onian low CFD-simula ion model and w o e he manusc ip . Sebas ian Wi h was in ol ed in he de ice ab ica ion and helped wi h he expe imen s a he synch o on beamlines a which Adeline Bu e , Gonzalo San o o, Shun Yu, Jan Pe lich, S ephan V. Ro h p o ided he echnical suppo . And eas Ma k was in ol ed in he µPIV-expe imen s and hei analysis. Ch is oph Hanske p o ided he ace pa icles o he µPIV-expe imen s. S ephan Fö s e supe ised he p ojec and co ec ed he manusc ip . 99 Chap e 7.3 This wo k is p epa ed as a manusc ip ha is eady o submission; en i led: "Mic o luidic Liquid Je Sys em wi h compa ibili y o a mosphe ic and high- acuum condi ions". by Ma in T ebbin, Kilian K üge , Daniel DePon e, S ephan V. Ro h, Hen y N. Chapman, and S ephan Fö s e *. I pe o med mos o he expe imen s & da a analysis, c ea ed he non-linea wo-phase low CFD-simula ion model and w o e he manusc ip . Kilian K üge was in ol ed in he de ice ab ica ion, scien i ic discussions and helped wi h he expe imen s. Daniel DePon e, S ephan V. Ro h and Hen y N. Chapman we e in ol ed in scien i ic discussions and helped wi h he co ec ion o he manusc ip . Fu he , Daniel DePon e was in ol ed in he ESEM- expe imen s. S ephan Fö s e supe ised he p ojec and co ec ed he manusc ip . Chap e 7.4 This pape is published in So Ma e 2012, 8, 12124; en i led: "Lyo opic phase beha io o polyme -coa ed i on oxide nanopa icles". by Sa a Mehdizadeh Tahe i*, S e en Fische , Ma in T ebbin, Sebas ian Wi h, Jan H. Sch öde , Jan Pe lich, S ephan V. Ro h and S ephan Fö s e *. I designed and ab ica ed he g id-based s a ic mic o luidic de ice using so li hog aphy. Sa a Mehdizadeh Tahe i pe o med mos o he expe imen s & da a analysis and w o e he manusc ip wi h S ephan Fö s e . Fu he , I helped, oge he wi h Sebas ian Wi h, Jan H. Sch öde wi h he expe imen s a he synch o on beamlines a which Jan Pe lich, S ephan V. Ro h p o ided he echnical suppo . S e en Fische was in ol ed in he syn hesis o polyme s and scien i ic discussions. S ephan Fö s e supe ised he p ojec and co ec ed he manusc ip . Chap e 7.5 This pape is published in Lab Chip, 2011, 11, 2362 and ea u ed in Chemis y Wo ld; en i led: "Ea ly de elopmen d ug o mula ion on a chip: Fab ica ion o nanopa icles using a mic o luidic sp ay d ye ". by Julian Thiele, Maike Windbe gs, Adam R. Aba e, Ma in T ebbin, Ho Cheung Shum, S ephan Fö s e , and Da id A. Wei z*. I de eloped he non-linea FEM-based CFD-simula ion model and con ibu ed o he simula ion- ela ed w i ing o he manusc ip . Julian Thiele pe o med mos o he expe imen s and w o e he manusc ip . Adam Aba e was in ol ed in scien i ic discussions. Maike Windbe gs pe o med he sp ay expe imen s in bulk and was in ol ed in scien i ic discussions. Ho Cheung Shum conduc ed he SEM analysis o he d ug. S ephan Fö s e 100 co ec ed he manusc ip . Da id Wei z supe ised he p ojec . Pa s o his wo k ha e been submi ed o pa en ing. Chap e 7.6 This pape is published in Nano echnology 2011, 22, 305303; en i led: "F ees anding ilms o c osslinked gold nanopa icles p epa ed ia laye -by-laye spin- coa ing". by Hend ik Schlicke, Jan H. Sch öde , Ma in T ebbin, Alexey Pe o , Michael Ijeh, Ho s Welle and Tobias Vossmeye *. I designed and ab ica ed he mic os uc u es using compu e aided design and pho oli hpg aphy, was in ol ed in hei mic oscopy as well as he scien i ic discussion. Jan H. Sch öde pe o med mos o he expe imen s and w o e he manusc ip oge he wi h Tobias Vossmeye . Hend ik Schlicke and Alexey Pe o we e in ol ed in he expe imen s ela ed o he ilms and he scien i ic discussions. Michael Ijeh was in ol ed in ligand syn hesis. Ho s Welle co ec ed he manusc ip . Fu he , Tobias Vossmeye supe ised he p ojec . Chap e 7.7 This pape is published in Z. Phys. Chem. 2012, 226, 569–584; en i led: "Adso p ion o sphe ical polyelec oly e b ushes: om in e ac ions o su ace pa e ning". by Ch is oph Hanske, Johann E a h, Ch is in Küh , Ma in T ebbin, Ch is ian Schneide , Alexande Wi emann, and And eas Fe y*. I p oduced a specially designed s amp o mic o con ac p in ing using so li hog aphy. Ch is oph Hanske pe o med adso p ion expe imen s, he mic o con ac p in ing, analyzed hese expe imen s, was in ol ed in scien i ic discussions, and w o e he manusc ip . Johann E a h pe o med he AFM in e ac ion measu emen s, analyzed hese expe imen s, was in ol ed in scien i ic discussions, w o e pa s o he manusc ip , and co ec ed he manusc ip . Ch is in Küh syn hesized he SPBs. Ch is ian Schneide syn hesized he model pa icles o SPBs. Alexande Wi emann de eloped he syn hesis p o ocol o he SPBs, was in ol ed in scien i ic discussions, w o e pa s o he manusc ip , and helped co ec ing he manusc ip . And eas Fe y analyzed he esul s, helped wi h discussions, and co ec ed he manusc ip . 101 102 7 Publica ions 103 104 7.1 Aniso opic pa icles align pe pendicula o low- di ec ion in na ow mic ochannels Ma in T ebbin,1 Dagma S einhause ,2,3 Jan Pe lich,4 S ephan V. Ro h,4 Wal e Zimme mann,5 Julian Thiele,6* and S ephan Fö s e 1* 1) Physical Chemis y I, Uni e si y o Bay eu h, D-95447 Bay eu h, Ge many 2) Max-Planck-Ins i u e o Dynamics and Sel -O ganiza ion, D-37073 Gö ingen, Ge many 3) Deu sches Ins i u ü Kau schuk echnologie, D-30519 Hanno e , Ge many 4) HASYLAB/DESY, D-22607 Hambu g, Ge many 5) Theo e ical Physics I, Uni e si y o Bay eu h, D-95447 Bay eu h, Ge many 6) Radboud Uni e si y Nijmegen, Ins i u e o Molecules and Ma e ials, NL-6525 AJ, Nijmegen, The Ne he lands The low o ien a ion o aniso opic pa icles in na ow channels is o impo ance in many ields anging om he spinning and molding o ibe s o he low o cells and p o eins h ough hin capilla ies. I is commonly assumed ha aniso opic pa icles align pa allel o he low di ec ion. When lowing h ough na owed channel sec ions one expec s he inc eased low a e o imp o e he alignmen . He e we show by mic o ocus synch o on X- ay di ac ion and pola ized op ical mic oscopy o he i s ime ha , a e passing a na ow channel sec ion, aniso opic colloidal pa icles align pe pendicula o he low- di ec ion. We ind his o be a gene al beha iou o aniso opic colloids, also obse ed o disk-like pa icles. The pe pendicula alignmen is s able, ex ending h oughou he emaining pa downs eam he channel. We show by mic o pa icle image elocime y and ini e elemen compu a ional luid dynamic simula ions ha he pe pendicula o ien a ion is due o he eloci y ield ha ing la ge pe pendicula g adien s in he expansion zone a e he na ow sec ion. Shea - hinning, a ypical p ope y o aniso opic pa icles, p omo es pe pendicula ex ensional and o ien a ion. Ou disco e y has impo an consequences when conside ing he low o ien a ion o polyme s, micelles, ibe s, p o eins o cells h ough na ow channels, pipes o capilla y sec ions. An immedia e consequence o he p oduc ion o ibe s is he necessi y o ealignmen by ex ension in low di ec ion. Fo ib ous p o eins, eo ien a ion and s able plug- low a e likely mechanisms o p o ein coagula ion. 105 I eo ien a ion in pe pendicula di ec ion in a channel expansion zone is a gene al p ope y o aniso opic colloids, i should ha e i s cause in he hyd odynamic low pa e n. Using mic o pa icle image elocime y we de e mined he eloci y p o ile ),( yx in he na ow sec ion as shown in Fig. 4. En e ing he channel sec ion, he e is i s a con ac ion zone wi h plana ex ensional low in low (x-) di ec ion, ollowed by an expansion zone wi h plana ex ensional low in pe pendicula (y-) di ec ion. Fig. 4 A shows he pola ized op ical mic og aph o he channel sec ion indica ing zones wi h pa allel (blue) and pe pendicula (yellow) low o ien a ion. Fig. 4B shows he measu ed low eloci y, ob ained om elocime y measu emen s o added 3.3 µm diame e ace pa icles. As expec ed, he low elocime y is la ges in he na ow sec ion o he channel. Fig. 4 C shows he measu ed eloci y p o iles ac oss he channel a posi ion I be o e en e ing he con ac ion zone, and a posi ion III a e he expansion zone. The eloci y p o iles a e bo h non-pa abolic, a consequence o he shea - hinning, non-New onian low beha io o he micelles. Bo h a e hyd odynamically s able s a es, ye ha ing di e en eloci y p o iles. The p e- ape ing eloci y p o ile a posi ion I has a b oad, bu clea ly no iceable maximum, whe eas he pos - ape ing eloci y p o ile a posi ion III is comple ely la , indica ing plug low. The eloci y p o iles can be swi ched back and o h in subsequen na ow channels as shown in Fig. S5 (Suppo ing In o ma ion). Fig. 4 D shows he measu ed eloci y componen s x and y ac oss he channel a posi ion II in he expansion zone. F om he eloci y componen s he shea a es xy ∇= γ ! and he ex ensional a es yy −∇= ε ! can be calcula ed and a e displayed in Fig. 4 E. We obse e ha in he middle pa o he c oss sec ion he ex ensional a e ε ! is ei he la ge , o a leas o compa able magni ude o he shea a e γ ! . Ex ensional lows a e much mo e e ec i e in o ien ing and aligning aniso opic pa icles compa ed o shea lows. [20,21] They lead o eo ien a ion i he ex ensional a es become compa able o he shea a es. Wi h an in e nal elaxa ion ime o he cylind ical micelles o 400≈ τ s as de e mined om he heological measu emen s and ex ensional a es o 2≈ ε ! s-1 (see Figu e 4 E), alues o he Debo ah numbe a e 1800 >>== ετ ! De . Unde hese condi ions he micelles a e highly suscep ible o low-induced alignmen . In Fig. 4 F we mapped egions whe e 14.0/ ≥ γε !! in yellow, and egions whe e 14.0/ < γε !! in blue o compa ison wi h he low bi e ingence pa e n in Fig. 4 A. We no e ha he nea -ze o alues o y in he egions be o e he con ac ion zone and a e he expansion zone o he channel lead o some sca e o he da a. Ye , we obse e ha by choosing a h eshold o 14.0/ = γε !! , egions o high ex ensional a es in Fig. 4 F ag ee well wi h egions o pe pendicula o ien a ion in Figs. 3 C and 4 A. A he channel walls shea low domina es such ha 14.0/ < γε !! and micelles emain o ien ed in low-di ec ion as obse ed expe imen ally. The beginning o he sha p ise o he shea a e γ ! close o he channel wall (see Figu e 4 E) de ines a ela i ely sha p ansi ion wi h a s able in e ace be ween zones o pe pendicula and pa allel cylinde o ien a ion. 112 To dis inguish ea u es o he low pa e n speci ically ela ed o pa icle aniso opy om ea u es ela ed o jus channel geome y, we pe o med nume ical compu a ional luid dynamic (CFD) simula ions o calcula e shea a es and ex ensional a es in he con ac ion/expansion zone. The calcula ions we e done o New onian liquids, bu also o non-New onian, shea - hinning liquids such as solu ions o wo mlike micelles. Shea - hinning was accoun ed o by measu ing he shea - a e dependen iscosi y o he micella solu ion using a cone-pla e heome e and i ing he measu ed low cu e o he C oss equa ion (see Suppo ing In o ma ion). This equa ion well desc ibes he measu ed da a und se es o pa ame e ize he low cu e in e ms o i s high- and low-shea iscosi y, he elaxa ion ime and a powe -law exponen , which can be implemen ed in he CFD-simula ions. De ails o he simula ions and he expe imen a e desc ibed in he Supplemen a y In o ma ion. Figu e 5 A shows he calcula ed eloci y ield ),( yx o a shea - hinning solu ion in he con ac ion/expansion zone o a ypical channel geome y and low a e used in he expe imen s. Figu e 5. Veloci y ield ),( yx o cylind ical micelles calcula ed by CFD simula ions. Uppe panel (A): Calcula ed eloci y in he wide and na ow channel sec ion. (B): Calcula ed eloci y p o iles )(y x in he p e- ape ing zone I (!!!) and pos - ape ing zone III (!!!). (C): Veloci y p o iles )(y x (!!!) and )(y y (!!!) in he expansion zone indica ed in (A). (D) Shea a e )(y γ ! (!!!) and ex ensional a e )(y ε ! (!!!) in he expansion zone. (E) Map o he a io γε !! / in he wide and na ow channel sec ion. In he o ange egions 14.0/ > γε !! , whe eas in he blue egions 14.0/ < γε !! . The esul ing colo map shows good ag eemen wi h he pola ized op ical mic og aphs in Fig. 3A, he X- ay in ensi y map in Fig. 3 C, and he measu ed eloci y map in Fig. 4 F. 113 Fig. 5 B shows he calcula ed eloci y p o iles ac oss he channel a posi ion I be o e en e ing he con ac ion zone, and a posi ion III a e he expansion zone. The eloci y p o iles a e bo h non-pa abolic, wi h an almos cons an low eloci y in he cen al pa o he channel and a s ongly dec easing low eloci y close o he channel walls, a consequence o he shea - hinning, non-New onian low beha io . This is di e en o New onian luids whe e he low eloci y has a con inuously a ying pa abolic p o ile as shown by mic o pa icle image elocime y and CFD calcula ions in he Suppo ing In o ma ion. Figu e 5 C shows he calcula ed eloci y componen s x and y , and Figu e 5 D he co esponding shea a e xy ∇= γ ! and ex ensional a e yy −∇= ε ! along he line ac oss he expansion zone indica ed in Figu e 5 A. We obse e, as in he expe imen al da a in Fig. 4, ha o e he majo cen al pa o he c oss sec ion he ex ensional a e ε ! is la ge o a leas o he same o de o magni ude as he shea a e γ ! . Figu e 5 E shows he calcula ed a io γε !! / o e he con ac ion/expansion zone wi h a colo scale adjus ed such as yellow colo indica es he zone whe e 14.0/ > γε !! , whe eas blue colo indica es he zone wi h 14.0/ < γε !! . A compa ison wi h Figu e 2, 3A, 3C, and 4 A shows ha choosing a h eshold o 14.0/ = γε !! also in he calcula ions nea ly quan i a i ely ep oduces he yellow zone wi h pe pendicula alignmen and he blue zone wi h pa allel alignmen in ou expe imen . CFD simula ions show ha wi h dec easing low a e and diame e o he ape ed c oss-sec ion he a ea o he pe pendicula o ien ed zone inc eases, which is in ag eemen wi h ou expe imen al obse a ions. The calcula ions do no ep oduce he di e ence obse ed in he wo s able p e- and pos - ape ing eloci y p o iles shown in Fig. 4 C. To accoun o his di e ence he e ec o aniso opic colloids on he shea ield has o be modeled in mo e de ail, which in he p esen calcula ions has only indi ec ly been accoun ed o ia he esul ing shea - hinning beha io . In conclusion, we show ha cylind ical micelles o ien ei he pa allel o he low di ec ion o , a e passing h ough a na ow channel sec ion, pe pendicula o he low di ec ion. Bo h o ien a ions a e s able downs eam he channel. Expe imen s wi h cylind ical micelles o di e en ype and ecen li e a u e indica e ha he eo ien a ion in pe pendicula di ec ion is gene ally occu ing o aniso opic cylind ical and disk-like colloids. The pe pendicula o ien a ion is caused by he eloci y ield ha ing la ge pe pendicula g adien s in he expansion zone a e he na ow sec ion. Shea - hinning, a ypical p ope y o aniso opic pa icles, p omo es pe pendicula ex ensional and o ien a ion. This phenomenon has impo an implica ions when conside ing he low o ien a ion o polyme s, ibe s, p o eins o cells h ough na ow sec ions such as dies, molds o ape ed capilla ies. An immedia e consequence o he p oduc ion o ibe s is he necessi y o subsequen ly apply ex ensional o ces o e-align polyme s o ib ils in low-di ec ion o op imal ibe mechanical p ope ies. Fo ib ous p o eins eo ien a ion and s able plug- low a e mechanisms o p o ein o cell coagula ion wi h possible ela ions o h ombosis. [22] Cu en expe imen s indica e ha pe pendicula low-o ien a ion can be u ilized o o ien cylind ical micelles pe pendicula o su aces, which is o ele ance o applica ions in ol ing elec ical o he mal anspo pe pendicula o a su ace such as in hyb id sola cells. 114 Ma e ials and Me hods Fab ica ion o mic o luidic de ices: The mic ochannel mas e o he mic o luidic de ice was ab ica ed using op ical li hog aphy.[23] The mic ochannel ne wo k was designed in Au oCAD 2011 and p in ed on a mask oil wi h an UV-abso ben ink (Zi zmann GmbH). An in e se black-whi e image o he de ice design is shown in Figu e 1A. To pump luids in o he de ice, inle po s a e in e aced wi h ubing. Thei punch loca ion is su ounded wi h polygons ha sca e ligh , making i easy o see and accu a ely punch he co esponding PDMS eplica ha a e ab ica ed using so li hog aphy, as desc ibed in de ail in he supplemen al pa o his publica ion.[24,25] S able ubing in e aces a e an impo an p e equisi e o long- e m in-si u scanning expe imen s a he synch o on beam line. P epa a ion o cylind ical micelle solu ion: Poly(isop ene-b-polye hylene oxide) (PI110-PEG198, mean Mw 16,200 g mol-1) was p epa ed by sequen ial li ing anionic polyme iza ion, yielding a block copolyme wi h na ow polydispe si y MW/Mn = 1.02, whe e Mw and Mn a e he weigh - and numbe -a e aged molecula weigh s. The syn hesis and cha ac e iza ion o PI-PEG is desc ibed in de ail elsewhe e.[26] Poly(e hylenebu ylene-b-polye hylene oxide) (PEB39-PEO102, mean Mw 7,700 g mol-1, MW/Mn = 1.06) was ob ained om EVONIK and lyophilized be o e use. The d y polyme s we e dissol ed in Millipo e-quali y wa e wi h a esis i i y o 18.1 MΩ cm-1 be o e use. The solu ions we e homogenized using an Ul aTu ax T8 (IKA We ke GmbH) and s o ed o allow he copolyme o swell in he wa e o h ee weeks a oom empe a u e. Be o e he mic o luidic expe imen s, he solu ions a e il e ed h ough a PTFE il e wi h 5 µm po e size. De ice ope a ion a he beamline: The expe imen s we e pe o med a he beamline BW4 and P03 a HASYLAB/DESY. The mic o luidic de ice was connec ed o high p ecision sy inge pumps (Ce oni GmbH, Nemesys sys em) and posi ioned in he X- ay beam. A e collec ing he necessa y backg ound da a o an emp y mic ochannel, he sy inge pumps a e se o ypical low a es o 32.4 µL h-1 co esponding o a mean s eam eloci y o 360 µm s-1. A e 15 min. o equilib a ion ime, measu emen s along he low di ec ion a e pe o med wi h a mic o ocused X- ay beam a a wa eleng h o λ = 0.1381 nm. A bo h beamlines he beam was 20 µm in wid h and 30 µm in heigh . X- ay sca e ing pa e ns we e eco ded wi h s ep sizes o 70 µm a a dis ance o 3.128 m behind he mic o luidic de ice using a Pila us 300K de ec o (Dec is L d.) wi h a pixel size o 172 µm by 172 µm. The in eg a ion ime is 240 s. 115 Re e ences [1] Yang HH, Allen SR (2000) in Ad anced ibe spinning echnology, (Ed: Nakajima T), Woodhead Publ., Abing on, England, Ch. 6. [2] Cuculo JA, Ho e JF, Zhou Q (2001) in S uc u e Fo ma ion in Polyme ic Fibe s, (Ed: D. R. Salem), 1s edn., Hanse Ga dne Pubns., Ch. 3. [3] Bouxsein NF, Hi s LS, Li Y, Sa inya CR, Samah ZA, MacDonald NC, Pynn R (2004) Alignmen o ilamen ous p o eins and associa ed molecules h ough con inemen in mic ochannels. Appl Phys Le 85:5775. [4] P. Bu le P (1999) Shea induced s uc u es and ans o ma ions in complex luids. Cu Opin Colloid In e ace Sci 4:214. [5] Squi es TM, Quake SR (2005) Mic o luidics: luid physics a he nanoli e scale. Re Mod Phys 77:977. [6] S one HA, S oock AD, Ajda i A (2004) Enginee ing lows in small de ices: mic o luidics owa d a lab-on-a-chip. Annu Re Fluid Mech 36:381. [7] Ba e P, Faucon M, Lopez J, C is obal G, Des emau F, Dodge A, Guillo P, La al P, Masselon C, Salmon JB (2006) X- ay mic o ocussing combined wi h mic o luidics o on-chip X- ay sca e ing measu emen s. Lab Chip 6:494. [8] Ma in HP, B ooks NJ, Seddon JM, Te ill NJ, Luckham PF, Kowalski AJ, Cab al JT (2010) Complex luids unde mic o low p obed by SAXS: apid mic o ab ica ion and analysis. J Phys Con Se 247:012050. [9] Li S, Liu N, Chan-Pa k MB, Yan Y, Zhang Q (2007) Aligned single-walled ca bon nano ube pa e ns wi h nanoscale wid h, mic on-scale leng h and con ollable pi ch. Nano echnol 18:455302. [10] Hesse HC, Beck R, Ding C, Jones JB, Deek J, MacDonald NC, Li Y, Sa inya CR (2008) Di ec imaging o aligned neu o ilamen ne wo ks assembled using in si u dialysis in mic ochannels. Langmui 24:8397. [11] Rammensee S, Slo a U, Scheibel T, Bausch AR (2008) Assembly mechanism o ecombinan spide silk p o eins. P oc. Na l. Acad. Sci. U.S.A. 105:6590. [12] Sun B, Si inghaus H (2006) Su ace ension and luid low d i en sel -assembly o o de ed ZnO nano od ilms o high pe o mance ield e ec ansis o s. J Am Chem Soc 128:16231. [13] Dimalan a ET, Lim A, Runnheim R, Lame s C, Chu as C, Fo es DK, De Pablo JJ, G aham MD, Coppe smi h SN, Golds ein S, Schwa z DC (2004) A mic o luidic sys em o la ge DNA molecule a ays. Anal Chem 76:5293. [14] Ca es ME, Candau SJ (1990) S a ics and dynamics o wo m-like su ac an micelles. J Phys Cond Ma 2:6869. [15] Fö s e S, Kon ad M, Lindne P (2005) Shea hinning and o ien a ional o de ing o wo mlike micelles. Phys Re Le 94:017803. [16] Rich e ing W (2001) Rheology and shea induced s uc u es in su ac an solu ions. Cu Opin Colloid In e ace Sci 6:446. 116 [17] Wa on G, Michels B, S eye A, Schossele F (2004) Shea -induced demixing and shea - banding ins abili ies in dilu e iblock copolyme solu ions. Mac omolecules 37:2313. [18] Gao C, Kulka ni SD, Mo is JF, Gilch is JF (2010) Di ec in es iga ion o aniso opic suspension s uc u e in p essu e-d i en low. Phys Re E 81:041403. [19] C ome M, Cook LP, McKinley GH (2011) P essu e-d i en low o wo mlike micella solu ions in ec ilinea mic ochannels. J Non-New onian Fluid Mech 166:180. [20] Singh AP, Rey AD (1995) Compu e simula ion o dynamics and mo phology o disco ic mesophases in ex ensional lows. Liquid C ys als 18:219. [21] Oli ei a MSN, Al es MA, Pinho FT, McKinley GH (2007) New onian luid low h ough mic o ab ica ed hype bolic con ac ions. Exp Fluids 43:437. [22] Liu Q, Mi c D, Fu BM (2008) Mechanical mechanisms o h ombosis in in ac ben mic o essels o a mesen e y. J. Biomechanics 41:2726. [23] Nguyen NT, We eley S (2002), Fundamen als and Applica ions o Mic o luidics, 1s edn., A ech House, Ch. 3. [24] Xia Y, Whi esides GM (1998) So li hog aphy. Annu Re Ma e Sci 28:153. [25] Quake SR, Sche e A (2000) F om mic o- o nano ab ica ion wi h so ma e ials. Science 290:1536. [26] Fö s e S, K äme E (1999) Syn hesis o PB-PEO and PI-PEO block copolyme s wi h alkyl li hium ini ia o s and he phosphazene base -BuP4. Mac omolecules 32:2783. 117 118 Suppo ing In o ma ion Fab ica ion o Kap on-PDMS-Kap on mic o luidic de ices (Figu e S1): A nega i e pho o esis (SU-8 50, Mic ochem Co.) is spin-coa ed on o a silicon wa e . A mask aligne (Süss Mik o Tec) is used o impa he mic ochannel s uc u e in o he pho o esis . We op imize he mas e de ice ab ica ion o ob ain mic ochannels wi h a e y uni o m heigh o 100 µm. Al hough PDMS is widely applied o eplica e he mic ochannel mas e de ice using so li hog aphy, PDMS sca e s and abso bs X- ays. To ab ica e X- ay compa ible mic o luidic de ices in ol ing PDMS, we modi ied he con en ional ab ica ion p ocedu e, based on he wo k o E ans and Doo z.[S1, S2] A e pou ing PDMS p e-polyme (Sylga d 184, Dow Co ning) on he mas e de ice, excess p e-polyme is emo ed om he mas e de ice wi h a azo blade. The emaining PDMS is cu ed and a small piece o sel - adhesi e polyimide ape (DuPon ™ Kap on®) is used o co e he a ea o in e es o he mic ochannel ne wo k including he cu ed and ape ed mic ochannel sec ions. A second laye o PDMS is cu ed on o he p e ious laye s. The PDMS eplica is emo ed om he mas e de ice and inle po s a e punched in o he polyme using a biopsy punch needle (Ha is Uni-Co e™ 0.75 mm). The bo om o he de ice is sealed wi h Kap on ape and a window is cu in o he op PDMS laye . Thus, he mic ochannels in he a ea o in e es a e solely sealed wi h X- ay anspa en Kap on ape. Figu e S1. Fab ica ion o mic o luidic de ices wi h X- ay analysis capabili y based on PDMS. (A) Mas e de ice ab ica ion using pho oli hog aphy. (B) Fab ica ion o X- ay anspa en mic o luidic Kap on-PDMS-Kap on sandwich de ices. (B1) PDMS is pou ed on a mas e de ice, (B2) and excess PDMS is cu -o he mic ochannel s uc u e. (B3) Sel - adhesi e Kap on ape is used o seal he a ea o in e es , (B4) and a second laye o PDMS is g a ed on op. (B5) The PDMS eplica is peeled o he mas e de ice, and inle po s o luids a e added. (B6) The bo om is sealed wi h Kap on ape, (B7, B8) be o e a window is cu in o he op PDMS laye wi h simila dimensions as he Kap on window in B3. 119 Compu a ional Fluid Dynamics Simula ions (CFD): The luid dynamics calcula ions a e based on he Na ie -S okes equa ions assuming an incomp essible luid, i.e. ρ = cons . [S3], ρ∇∙!=0 ρ ∂! ∂!+ρ!∙∇!=∇∙−!!+!∇!+∇!!+! wi h he densi y o he luid ρ, he p essu e !, he iden i y ma ix !, he dynamic iscosi y o he luid !, he eloci y ield u and he olume o ce F. Solu ions o aniso opic pa icles exhibi p onounced shea - hinning. To model he esul ing low p o ile, we used he so wa e package COMSOL Mul iphysics 4.2a, which allows one o impo CAD-designed mic ochannel geome ies and akes in o accoun Non-New onian low beha io in compu a ional luid dynamics simula ions. [S3] To in eg a e non-linea low beha io in he CFD-model, expe imen al da a a e employed ob ained by heome y using a Bohlin Gemini 200 which was used o measu e he shea - a e dependence o he iscosi y. The measu ed low cu e can be well desc ibed by he C oss equa ion: [S4, S5] η=η!+η!−η! 1+(τ!γ)! whe e he iscosi y!η is desc ibed by he ze o-shea iscosi y η!, he high-shea iscosi y η!, he in e nal elaxa ion ime τ! and he powe -law exponen n cha ac e izing he shea hinning be ween η! and η!. This equa ions well ep oduces he expe imen al da a as shown in Fig. S2. The alues o he pa ame e s ob ained by i ing he equa ion o he measu ed low-cu e a e hen used in he CFD-calcula ions. Figu e S2. Expe imen al heological da a: iscosi y (η) as a unc ion o he shea a e (γ) is desc ibed by he C oss equa ion. The simula ions yield he eloci y ield ),( yx om which he shea - and ex ensional a es can be calcula ed. A mean shea a e [S3] 120 γ= 4u! !+2u!+ ! !+4 ! ! 2 wi h he co esponding x- o y-componen s o u ( eloci y in x-di ec ion) and ( eloci y in y- di ec ion) is used o map he local eloci y on he mic ochannel. The model is sol ed o 1412784 ini e elemen s and 1072212 deg ees o eedom using a mul i on al massi ely pa allel sol e (MUMPS). The a e age elemen quali y o he mesh is 0.9889 on a scale om 0 o 1, whe e 1 is he highes quali y; he minimal elemen quali y is 0.6923. Using a Windows 7 x64 machine wi h wo quad-co e In el® Xeon® E5440- p ocesso s ope a ing a 2.83 GHz and an in e nal memo y o 32 GB RAM. All ele an pa ame e s, which a e used in he simula ions, a e summa ized in Table S1. Table S1. Ma e ial p ope ies used in he simula ion model. Figu e 3 & S3 Non- New onian New onian Flow a e low 32.4 µL h-1 32.4 µL h-1 Flow speed 18.52 mm s-1 18.52 mm s-1 ze o-shea iscosi y η! 19522 Pa s - high-shea iscosi y η! 0.1 Pa s - in e nal elaxa ion ime τ! 388.5 s - powe -law exponen n 0.99 - Viscosi y η [S6] - 1.002∙10-3 kg m-1 s-1 Densi y ρ [S6] 998.2 kg m-3 998.2 kg m-3 Tempe a u e T 293.15 K 293.15 K Calcula ed eloci y ield o wa e The calcula ed eloci y ield o he wo mlike micelles has been discussed in he main pa o he publica ion. He e we p esen he same calcula ions o wa e as a New onian luid. This p o ides insigh o wha ex en he shea - hinning beha io causes he obse ed eo ien a ion beha io . We calcula ed he eloci y ield o wa e unde he same expe imen al condi ions as o wo mlike micelles. The esul s a e shown in Fig. S3, which is o ganized simila o Figu e 5 in he main sec ion o he manusc ip o compa ison. 121 In oduc ion Mic o luidic SAXS expe imen . Mic o luidics enables he p ecise con ol o liquids on he nanoli e scale. 1 These e y well de ined low condi ions make his echnology p edes ined o undamen al in es iga ions a mic o ocused X- ay sou ces. The basic idea o his expe imen al se up is o u ilize he e y well de ined con inuous low condi ions o he mic ochannel o scan i wi h a X- ay mic obeam and eco d he small-angle X- ay sca e ing (SAXS) pa e n o each measu ed posi ion. This mapped da a hen allows o ge a as and de ailed o e iew o e he low expe imen wi hin he mic ochannel. The combina ion o mic obeam X- ay sca e ing and mic o luidics is cu en ly being de eloped in o a powe ul expe imen al me hodology sui able o he in es iga ion o nanos uc u es, pa icle alignmen and he in si u s udy o kine ics by c ea ing X- ay compa ible mic o low chips and mic o luidic liquid je de ices. 2-10 The sample is pumped h ough he mic ochannels o a X- ay compa ible de ice and he mic o ocused X- ay beam passes he lowing sample; in his case aniso opic wo mlike micelles. The sca e ed X- ays, which con ain he s uc u al in o ma ion, a e eco ded using a 2D digi al de ec o (Pil a us 1M, Dec is). The ypical expe imen al se up a he mic o ocus beamline P03 (PETRA III, DESY) is shown in he supplemen al in o ma ion in Fig.S1. 11 Recen ly, mic o luidic SAXS scanning expe imen s a he P03 and BW4 beamlines (DESY, Hambu g) e ealed he s iking e ec , ha a e passing a na ow sec ion, wo mlike pa icles a e o a ed pe pendicula o he low di ec ion, keeping his o ien a ion o e he emaining leng h o he channel. 2 The low-alignmen o cylind ical, wo mlike o ib ous s uc u es is cen al o many p ocessing s eps such as in he p oduc ion o ibe s, du ing injec ion molding o he low o cells and p o eins h ough hin capilla ies. 12-16 In his pape , his pe pendicula o ien a ion will be in es iga ed in mo e de ail o unde s and he in luence o expe imen al pa ame e s on he o ien a ion e ec . Fo his ask, we c ea e X- ay compa ible mic o luidic de ices which a e made o NOA81 (No dland Op ical Adhesi e 81) o he SAXS analysis. 9,17-19 A edesigned NOA81- ab ica ion ou ine now allows he p oduc ion o e y hin de ices o high X- ay ansmission while also main aining he ma e ial’s e y good sol en compa ibili y. The esul ing mic o luidic de ice is shown in Fig.1B while i s de ailed ab ica ion ou ine is explained in he expe imen al sec ion. This ou ine is so li hog aphy-based because i uses in e ed PDMS-mic ochannels as a molding empla e. 20,21 The e o e, he high mic o luidic design lexibili y is main ained which is impo an o apid p o o yping. 1 Wi h his gi en design con ol om so li hog aphy, we a y he geome ic pa ame e s o he mic ochannel like channel wid h (250 o 500 µm) ape ing a ios (1:10 o 1:2.5; w. . . channel wid h) and -leng hs (9:1 o 2:1; w. . . channel wid h), as illus a ed in Fig.1A. Fu he , we a y he sample concen a ion (30 o 0% w/w) and low speeds (100 o 2000 µl h-1). 128 Figu e 1 (A) The CAD-based design and use o high esolu ion pho o masks (128 kdpi, JD Pho o, UK) enable p ecise con ol o e he mic ochannel geome ies. In his case he shea ield is con olled by smoo h ape ings wi h a ying a ios (1:10 o 1:2.5) and leng hs (9:1 o 2:1) wi h espec o he o iginal channel wid hs (250 o 500 µm) a a cons an channel heigh o 100 µm. Fu he , he a ied ele an expe imen al pa ame e s include low speed and sample concen a ion. (B) The esul ing NOA81-de ice and a ypical elec on mic og aph o he SU-8 eplica ion empla e ( ape ing a io 3:1, channel wid h 150 µm). Resul s and Discussion SAXS pa e n analysis. The mic o luidic SAXS scanning se up enables as sample sc eening o he a ying expe imen al condi ions. Each ob ained SAXS pa e n o a gi en posi ion con ains a ious s uc u al in o ma ion abou he cylind ical micelles, like i.e. pa icle size, uni cell dimensions and o ien a ional dis ibu ion; an example SAXS pa e n is shown in Fig.2. The e ically-posi ioned c escen s in his aniso opic SAXS pa e n indica e a pa allel o ien a ion o wo mlike micelles in espec o he (ho izon al) low di ec ion. Addi ionally, he azimu hal peak wid h con ains in o ma ion abou he micelles’ o ien a ion dis ibu ion. A e he adial a e aging o his SAXS pa e n, he i s and p ojec ions can be calcula ed using he analysis so wa e Sca e which yields u he s uc u al in o ma ion abou he sample. 22,23 This sample con ains hexagonally closes packed micelles (PEB39-b-PEO102, 30% w/w in pu e wa e ) which a e o ien ed pa allel o he (ho izon al) low di ec ion. Thei adius is 10.0 nm wi h a ying leng hs up o he mic on ange and a uni cell size o 40.9 nm (1 nm displacemen , Lague e dis ibu ion wi h =14). 23 This so wa e also allows he calcula ion o he micelle’s 2D sca e ing pa e ns based on he 3D model o hexagonally closes -packed cylinde s as shown in he image inlay in Fig.2. 23 The i ing pa ame e s a e lis ed in he ollowing Tab.1. 129 Table 1 Lis o i ing pa ame e s o he SAXS-pa e n calcula ion using he analysis so wa e Sca e . 22 Pa ame e Value Model Hexagonally packed cylinde s (P6/mm) Micelle adius , nm 10.0 Rela i e s anda d dis ibu ion , nm 0.15 Cylinde leng h , nm 73.2 Rela i e s anda d dis ibu ion , nm 0.1 Uni cell dimensions , nm 40.9 Displacemen , nm 1.0 Radial domain size , nm 213 Radial domain size , nm 65 Dis ibu ion unc ion ype Lague e Dis ibu ion unc ion pa ame e 14 Figu e 2 The ob ained SAXS pa e ns con ain a ious s uc u al in o ma ion o he cylind ical micelles, like i.e. pa icle size ( =10 nm), uni cell dimensions (40.9 nm) and o ien a ional dis ibu ion (Lague e dis ibu ion wi h =14). The e ical c escen s in his aniso opic pa e n indica e pa allel o ien a ion in espec o he low ho izon al di ec ion. A e he adial a e aging o his SAXS pa e n, he i s and p ojec ions a e calcula ed using he analysis so wa e Sca e . 22,23 Fu he , i is also possible o calcula e he 2D sca e ing pa e n, as shown in he bo om igh co ne , based on he co esponding 3D model o hese hexagonally closes -packed cylinde s. 130 Small-angle X- ay sca e ing is a complemen a y me hod ha can be co ela ed o he esul s om o he echniques such as mic o pa icle image elocime y (µPIV), pola iza ion mic oscopy and CFD-simula ions which will be discussed o e he cou se o his pape . Gene a ion o colo -coded pixel maps om SAXS da a. A as o e iew o e he ob ained SAXS pa e ns o a single expe imen is gained by he eal- ime gene a ion o pixel maps based on he SAXS scanning loca ions. The SAXS pa e n’s s uc u al in o ma ion is hen used o colo -code he indi idual pixels based on he a e aged in ensi y wi hin micelle- o ien a ion dependen egions o in e es (ROI). The esul ing pixel maps o di e en ROIs a e shown in Fig.3B. The i s pixel map shows he a e aged in ensi y o all micelles o ien a ions and gi es an imp ession abou he mic ochannel’s shape. The ROI o he micelle o ien a ion pa allel o he low is shown below and he colo -coded pixel maps a e in good ag eemen wi h he images ob ained om pola iza ion mic oscopy (Fig.3A). In he SAXS- based pixelmap we obse e ha he pa allel o ien a ion is s onges in he beginning o he ape ing, whe e bo h ex ensional and shea o ces a e applied in low di ec ion. Pa allel micelle o ien a ion is also ound close o he channel walls whe e he wall shea is mos dominan due o he low o he luid. The hi d ROI-based pixel map ep esen s he a eas whe e he micelles a e o a ed pe pendicula o he low due o he dominance o ex ensional o ces which a e di ec ed e ically o he main low di ec ion. 2 This pe pendicula o ien a ion can be also obse ed well as an o ange-colo ed a ea in pola iza ion mic oscopy o which he SAXS scanning posi ion can be co ela ed as indica ed in Fig.3A. Las ly, he plus 45° and minus 45° micelle o ien a ions a e shown below and a e ei he he esul o pa allel micelle alignmen o he il ed mic ochannel walls o he ape ing o due o he ansi ion om pa allel o pe pendicula o ien a ion. The e ec o he ape ing a io on he pe pendicula o ien a ion. The eloci y low ield in he mic ochannel and, hence, he shea and ex ensional o ces a e a ied by changing he ape ing a ios o he mic ochannel’s na ow sec ion be ween 10:1 and 2.5:1 wi h espec o he mic ochannel wid h. The low o micelles h ough na ow sec ions unde he same condi ions is s udied using pola iza ion mic oscopy, as shown in Fig.4A. We ind ha he pe pendicula o ien a ion, which is indica ed by he o ange egion, has he b oades wid h and is mos p onounced a he la ges ape ing aspec a io o 10:1. Bo h he colo -in ensi y and he wid h o he o ange s eam dec ease wi h wide mic ochannel ape ings. The same se o expe imen s is also s udied using mic o luidic SAXS o which he esul ing colo -coded pixel maps a e shown in Fig.4B,C. These pixel maps sha e he same colo -scale o illus a e he ela i e in ensi y o e-o ien a ion. Simila o he obse ed ends using pola iza ion mic oscopy, we ind ha he e-o ien a ion inc eases wi h la ge ape ing a ios and ha less pe pendicula ly o ien a ed micelles a e obse ed wi h lowe he ape ing aspec a ios. 131 Figu e 3 (A) Pola iza ion mic oscopic (PM) images o he se up using c ossed pola ize s wi h a qua e wa e pla e gi e an imp ession abou he lowing micelle’s o ien a ion: he blue colo in he PM co esponds o a pa allel o ien a ion, while o ange a eas co espond o a pe pendicula o ien a ion o micelles in espec o he low di ec ion. The channel wid h is 500 µm wi h ape ing na owing down o 50 µm (10:1) o e a leng h o 3500 µm (7:1). The micelle concen a ion is 30% w/w a a low speed o 200 µl h-1. (B) This o ien a ion is con i med by he SAXS expe imen s a he P03/MiNaXS beamline which can easily be co ela ed wi h he PM because he condi ions, such as low speed, channel wid h, ape ing a io and leng h, e c. a e highly ep oducible in he mic o luidic de ice. The lowing complex luid is scanned a a gi en se o posi ions using he mic o- ocused X- ay beam. Each ob ained pixel co esponds o a sca e ing pa e n and he e o e con ains he ull s uc u al in o ma ion. (C) Colo -coding hese pixels based on he a e aged in ensi y a a gi en egion o in e es (ROI) gi es a de ailed o e iew o e he s uc u al e olu ion. The basic ypes o o ien a ional dis ibu ion in espec o he low di ec ion a e iso opic, pa allel, pe pendicula , il ed plus 45° o minus 45°. The colo ba s o hese pixel maps is adjus ed indi idually o exp ess he o ien a ion egions mo e clea ly. 132 Figu e 4 Compa ison o he in luence o he ape ing a ios on he micelle o ien a ion beha io . (A) Pola iza ion mic oscopic images unde he same condi ions as he ollowing SAXS pixel maps (micelle concen a ion 30% w/w, 200 µl h-1) which show he egions o pa allel (B) and pe pendicula (C) o ien a ion a di e en ape ing a ios (10:1, 5:1, 2.5:1) wi h espec o he channel wid h o 500 µm. The colo -scale is equal o all pixel maps o illus a e he ela i e in ensi y o e-o ien a ion. 133 Analysis o he in luence o he ape ing a io on he o ien a ion dis ibu ion. The o ien a ion dis ibu ion change along he mic ochannel has o be conside ed o he de ailed analysis o he in luence o ape ing a ios on he in ensi y o pe pendicula micelle o ien a ion. This o ien a ion in o ma ion can be ex ac ed om a se ies o SAXS pa e ns om he ape ing a io a ia ion (Fig.4B,C). These e ical scans wi h mic o ocused X- ay beams, along he mic ochannel o each ape ing aspec a io, a e compa ed in he Fig.5 p o iding a i s quali a i e analysis. He e, he i s and las ows o SAXS pa e ns show he pa allel micelle o ien a ion close o he walls which is indica ed by he e ical posi ion o he peaks. The second and ou h ows show ansi ion egions whe e he wo mlike micelles change om pa allel o pe pendicula o ien a ion and ice e sa. This ealignmen appea s as ‘ ish’- shaped o c escen -like peaks in he SAXS pa e ns ha o igina e om an asymme ical o ien a ional dis ibu ion along he azimu h. The scanning esolu ion o his ansi ion is de e mined by he size o he X- ay mic obeam which is 20x30 µm2 in ou expe imen s. The e o e, he obse ed ‘ ish’-shape could o igina e om a much sha pe ansi ion egion ha is smea ed by he o e laying X- ay beam o wo scanning posi ions a pa allel and pe pendicula micelle o ien a ion. Howe e , he colo ansi ion in pola iza ion mic oscopic images sugges s a smoo h o ien a ion ansi ion. A smalle mic o ocused X- ay beam could be used o esol e his egion in mo e de ail and s udy his ansi ional laye in mo e de ail. The cen al pa o he mic ochannel scan is shown in he hi d ow. We ind ha he micelle o ien a ion dis ibu ion s ongly depends on he ape ing aspec a io. In case o he 10:1 a io, almos all micelles a e aligned pe pendicula ly o he low which is indica ed by he peak in ho izon al posi ion. Howe e , he ho izon al peak in ensi y and, hence, pe pendicula micelle o ien a ion dec eases wi h lowe ape ing a ios. In case o he 2.5:1 a io, he SAXS pa e n shows a mo e iso opic o ien a ion cha ac e is ic wi h a p e e ed pa allel o ien a ion. 134 Figu e 5 The de ailed SAXS analysis is based on a e ical line scan ac oss he mic ochannel a he ape ing exi , as indica ed by he long black a ow in he pola iza ion mic oscopic image a he op. Below, he SAXS-pa e ns o womlike micelles in pa allel and pe pendicula o ien a ions as well as he ansi ional s age a e shown. He e, he a io dec eases wi h each column om le (10:1) o igh (2.5:1). The ed box ma ks he SAXS-pa e ns wi h a e analyzed in mo e de ail in Fig.6. 135 Nex , his end can be analyzed quan i a i ely by choosing a cake-like ROI o he SAXS pa e ns and calcula ing he azimu hally a e aged in ensi y as indica ed in Fig.6A. This a e aging esul s in angle-dependen in ensi y cu es which con ain he peak in ensi ies, - wid hs and - o a ion angles o he pa allel and pe pendicula micelle o ien a ions a he speci ic scan posi ion in he mic ochannel. These cu es a he cen al scan posi ions (see ma k in Fig.6A) o all h ee ape ings a e combined in a single g aph (Fig.6C) which gi es an o e iew o e he peak in ensi y changes. This g aph e eals ha he o e all measu ed in ensi y s ays cons an and ha only he o ien a ion dis ibu ion changes wi h he ape ing a io. The peak o he pa allel o ien a ion a his cen al scan posi ion (Fig.6C) dec eases wi h a aising ape ing a io while he pe pendicula o ien a ion peak aises acco dingly. Fo a be e gene al o e iew, his ex ac ion p ocess o angle-dependen in ensi y cu es (see Fig.6C) is pe o med o he comple e se ies o SAXS pa e ns. The esul s o he h ee e ical scans a e hen combined in indi idual 2D plo s which a e shown in Fig.6B whe e he colo ep esen s he ex ac ed peak in ensi ies. These 2D colo plo s now allow o di e en ia e be ween wo egions o o ien a ion in ensi y which a e based on he pa allel o pe pendicula micelle o ien a ion. Simila o he desc ibed o ien a ion ends om abo e, we ind ha he ela i e peak in ensi ies change wi h he ape ing a io whe e he mos p onounced pe pendicula o ien a ion is ound in he cen al pa o he channel wi h a ape ing a io o 10:1. In case o he 5:1 a io, he non-pa allel o ien a ion is weake , bu s ill clea ly isible and has abou he same in ensi y le el as he pa allel o ien a ion. This end con inues wi h he 2.5:1 ape ing a io whe e he non-pa allel o ien a ion me ely e eals i sel as an iso opic o ien a ion dis ibu ion wi h a p e e ed pa allel o ien a ion. These micelle e-o ien a ion shi s can also be quan i ied by he in ensi y a io which we de ine as This a io desc ibes he ela ion be ween he summed peak in ensi ies o micelles wi h pe pendicula o ien a ion ( om 61 o 117°) wi h he combined peak in ensi y sums o pa allel and pe pendicula micelles ( wi h om −37 o 40°). This measu e allows o co ela e he ela i e peak in ensi ies om he SAXS expe imen s wi h he o ange/blue- con as obse ed in pola iza ion mic oscopy. We ind ha a alue o = 0.1 co esponds o he blue-colo ed egions wi h a p e e ed pa allel micelle o ien a ion, i.e. in p oximi y o he channel walls. The weakly o ange-colo ed egion a low ape ing a ios (2.5:1, see Fig.4A) can be obse ed a = 0.25 which co esponds o an iso opic micelle dis ibu ion. Regions wi h a mo e p onounced o ange colo which can be ound a highe ape ing a ios (5:1 & 10:1, see Fig.4A) indica e ha he o ien a ion dis ibu ion is domina ed by a pe pendicula micelle alignmen wi h - alues > 0.35. The obse ed ends can also be e i ied by mic opa icle image elocime y (µPIV) in he ollowing sec ion o his pape . 136 Figu e 6 Ex ended o e iew o e he micelle o ien a ion dis ibu ion analysis om he SAXS pixel maps in Fig.4B,C. (A) By azimu hally a e aging a cake-like slice om he SAXS image, i is possible o s udy and quan i y he micelle’s o ien a ion dis ibu ion. (C) The esul ing angle-dependen peak in ensi y cu es e eal he in luence o he mic ochannel’s ape ing a io on he o ien a ion dis ibu ion. (B) This e ec can be s udied in mo e de ail by ex ac ing he azimu hally a e aged cu es o all posi ions ac oss he mic ochannel di ec ly a e he ape ing and o di e en ape ing a ios. The esul s a e combined in 2D colo maps o he angle-dependen peak in ensi y which gi e an o e iew o e he micelle o ien a ion changes ac oss he mic ochannel ha a e quan i ied by he in ensi y a io . 137 o wo mlike micelles, nume ical CFD-simula ions ha a e based on expe imen al heome ic da a a e in good ag eemen wi h he esul s om small-angle X- ay sca e ing, pola iza ion mic oscopy and mic opa icle elocime y. 2 Since he cu en sys em is based on he same block copolyme (PEB39-PEO102, E onik), we assume he CFD-simula ions o be also alid o p edic ing he gene al ends o he luid’s low when he heome ic sample pa ame e s a e a ied heo e ically. The iscosi y change o he micelle solu ion is desc ibed by he Cole- Cole- (o C oss-) equa ion which is gi en in he expe imen al sec ion o his pape . F om his o mula, he ze o shea iscosi y as well as he in e nal elaxa ion ime a e a ied in a se ies o CFD-simula ions. Fo his, h ee di e en alues o each pa ame e a e chosen while all o he pa ame e s a e kep cons an , as summa ized in Tab.1. An o e iew o e he heological cu es ha esul om hese pa ame e combina ions a e shown in Fig.10A. The di e en colo s co espond o he a ying ze o shea iscosi ies while he symbol shapes change wi h a ying in e nal elaxa ion imes. These a ia ions a e o ep esen he heological changes o a sample wi h di e en concen a ions o ypes o aniso opic pa icles o s udy he gene al ends. Al hough simila heological p ope y-changes could also be achie ed and s udied expe imen ally, his CFD-simula ion-based app oach is easie o con ol and less ime-consuming. These di e en heological cu es o de ined pa ame e a ia ions a e hen used in he CFD-model o desc ibe he luid’s heome ic p ope ies. 2 Nex , he esul ing eloci y low ields om he simula ion (no shown) a e used o calcula e he shea and ex ensional a es. As shown by he µPIV-expe imen s and pola iza ion mic oscopy, he egions o pe pendicula micelle o ien a ion can be ound whe e he ex ensional a e is compa able o he shea a e .2 The e o e, his a io ( ) is used o colo -code he simula ed luid low inside he mic ochannel ape ings. The esul ing 2D colo maps a e a anged acco ding o he pa ame e changes and shown in Fig.10B. An o ange colo in hese 2D plo s co esponds o a egion whe e he micelle o ien a ion is ypically ound o be pe pendicula exceeding a alue o 0.14. Acco dingly, he blue a ea ep esen s he egions o p e e ed pa allel o ien a ion wi h a a io below 0.14. The simula ed esul s clea ly show ha he sample’s ze o shea iscosi y has a s ong in luence on he wid h o he o ange egion a e he ape ing. The highe he ini ial iscosi y o he sample, he mo e p onounced is he plug-like low o he sample and he g ea e is he shielding o he cen al micelles agains he wall-induced shea . The in e nal elaxa ion ime con ols he capabili y o he micelles in he cen e o change hei o ien a ion while lowing h ough he ape ing. Since his is a con inuously lowing sample, he geome ic shea ield o he ape ing can be seen as a e-o ien a ion impulse. The sho e he in e nal elaxa ion ime, he mo e a e he aniso opic pa icles capable o eac ing o his impulse and change hei o ien a ion acco ding o he s onges o ien ing o ce. Fu he , he sho e he in e nal elaxa ion ime, he less a e he aniso opic pa icles a ec ed 144 by wall shea e ec s because hei iscosi y s a s o dec ease a highe shea a es. Consequen ly, heological p ope ies ha a e ypical o non-New onian samples, like i.e. shea hinning and plug-like low, s ongly inc eases he endency o ind pe pendicula ly o ien ed micelles in mic ochannel a e con ined geome ies. This obse a ion is in ag eemen wi h he abo e-discussed esul s om small-angle X- ay sca e ing, pola iza ion mic oscopy and mic opa icle image elocime y. Figu e 10 CFD-simula ions based on heo e ical a ia ion o heological sample pa ame e s. (A) Cu es a e calcula ed based on he Cole-Cole-equa ion (see expe imen al sec ion) ha desc ibe he luid’s iscosi y change unde he in luence o shea . (B) CFD-simula ion esul s ha show colo plo s o he a io ( ) be ween he ex ensional a e and he shea a e . The o ange colo ep esen s he egions whe e he micelle o ien a ion is ypically ound o be pe pendicula (SAXS, µPIV, pola iza ion mic oscopy) because he men ioned a io exceeds a alue o 0.14. Acco dingly, he blue a ea ep esen s he egions o p e e ed pa allel o ien a ion wi h a a io below 0.14. 145 Expe imen al P epa a ion o Cylind ical Micelle Solu ion. The o ma ion o wo mlike micelles is based on he sel -assembly o he amphiphile block copolyme PEB39-PEO102 (lyophilized be o e use, mean Mw = 7,700 g mol-1, Mw/Mn = 1.06, E onik). 25-27 The polyme is dissol ed in pu e wa e (Millipo e-quali y wi h a esis i i y o 18.1 MΩ cm-1), homogenized using an Ul aTu ax T8 (IKA We ke GmbH) and s o ed a oom empe a u e o h ee weeks o allow he copolyme o swell. P io o he use in mic o luidic de ices he solu ion was il e ed h ough a poly e a luo e hylene il e wi h 5 µm po e size. P epa a ion o monodispe e polys y ene (PS) ace sphe es o µPIV. The µPIV- expe imen s equi e o add small sphe ical ace pa icles o he samples (4% w/w). The dispe sion polyme iza ion syn hesis and cha ac e iza ion o hese monodispe se polys y ene sphe es (1.65 µm adius, PDI 1.01) is desc ibed in de ail elsewhe e. 28 To a oid clogging o he PS-sphe es o he PDMS-mic ochannel walls, he pa icle hyd ophilici y is inc eased by a ea men wi h polyac ylic acid (PAA, 1 g/l in 100mM NaCl) o 15 min unde con inuous shaking. 28 The excess o PAA was emo ed by e-suspending he pa icles h ee imes wi h pu e wa e (Millipo e quali y). Mic o luidic (SAXS) de ice ab ica ion. The mic ochannels a e designed using Au oCAD 2013 (Au odesk) and p in ed on a high- esolu ion emulsion ilm mask (JD Ph o o). An example o he channel geome y is shown in Fig.1A. The subsequen ab ica ion o mas e - empla es o eplica ion and PDMS-de ices (sealed wi h glass-slides) o pola iza ion mic oscopy and he µPIV-expe imen s a e based on s anda d so li hog aphy which is desc ibed elsewhe e. 20,21 This p ocess in ol es spin coa ing he pho o esis SU-8 50 (Mic ochem) on o polished silicon wa e s (Si-Ma ) o con olled channel heigh s (100 µm) and is illus a ed in Fig.S41. The ab ica ion o he SAXS-compa ible NOA81-de ices includes an addi ional mic os uc u ing s ep using a mask aligne (Süss Mik o Tec) o he gene a ion o a wo-laye mic os uc u e as poin ed ou in Fig.S42. The i s SU-8-laye enables o con ol he window hickness by adding an addi ional gap laye nex o he mic ochannels o he expe imen s. The second laye con ains he mic ochannels o he liquid low (and he second laye o he gap s uc u e). The inal SU-8-mas e empla e is hen eplica ed using s anda d PDMS-based so li hog aphy using Sylga d 184 (Dow Co ning) as illus a ed in Fig.S43. 20,21 Nex , he PDMS- eplica is used o mic os uc u ing he UV-cu able adhesi e NOA81 (No dland). 17,18 This me hod is based on a NOA-PDMS-hyb id ou ine ha is desc ibed elsewhe e, 9 bu i is modi ied o allow an imp o ed sol en compa ibili y. 18 Ins ead o d illing holes h ough he inished de ice and sealing i wi h PDMS, glue and apes, a di e en app oach was used. 146 Fi s , holes a e punched in o he eplica ed PDMS mic ochannel- empla e using a biopsy needle (Ha is Uni-Co e). Nex , he ubings which only se e as a empla e, a e inse ed in o hese holes be o e he liquid NOA81 is pou ed o e he PDMS- empla e. A la PDMS-block wi h he in ended scan-window size is laid o e a liquid NOA81-d op which sp eads i sel . The wo-laye ed gap s uc u e nex o he mic ochannels con ol he de ice ma e ial heigh in his s ep, while he liquid NOA81-d op nex o he empla e- ubings should ha e a heigh o ca. 2-3 mm. A e UV-cu ing, he PDMS-block and he empla e ubings a e emo ed be o e he open mic ochannels a e sealed wi h a hin NOA81- ilm which was p epa ed in a simila ashion: no empla e- ubing, gap s uc u es and la (o al e na i ely 3D-mic ochannel- s uc u ed). The sealing o he de ice wi h a second NOA81- ilm by UV-cu ing is possible because he gas-pe meabili y a he PDMS-NOA81-in e ace inhibi s he comple e cu ing. 9 This inhibi ion lea es a hin cu able laye o he de ice sealing a he NOA81-su ace. The esul ing NOA81-based mic o luidic de ice is e y hin in he X- ay mic obeam scanning a ea (ca. 250 µm) wi h a homogenous channel heigh and he mic o luidic de ice has a e y low X- ay backg ound signal. Fu he , he inc eased ma e ial hickness a he inle s, which we e o med by he empla e- ubings, allow o a ach he ubings di ec ly o he de ice wi hou using an addi ional ubing-in e ace PDMS-laye as desc ibed ea lie . 9 The e o e he ull sol en compa ibili y o NOA81 (i.e. wa e , unpola o ganic sol en s, e c.) is main ained. 18 Fo compa ibili y wi h highe p essu es and o gua an ee eliable, s able ubing in e aces (i.e. o p ecious synch o on beam imes), i is sugges ed o glue he ubings o he de ice using wo-componen epoxy glues (Loc i e, Henkel). Mic oscopic se ups. The pola iza ion mic oscopic expe imen s a e eco ded using he in e ed mic oscope Axio e S100 (Zeiss). The o ange and blue colo ep esen a ion o igina es om using c ossed pola ize s and a qua e wa e pla e. The images a e eco ded using a Nikon D7000 came a. The mic opa icle image elocime y se up in ol es an IX71 in e ed mic oscope (Olympus) wi h a 100 W ungs en ligh ha was ocused on o he senso o a Phan om 9.1 (Vision Resea ch) high speed came a (2 µs minimum exposu e, maximum a e o 2 Gpx s-1). The eco ded ideo ames a e analyzed using he open sou ce so wa e package JPIV which gene a es eloci y ields by co ela ing he cap u ed ames. 24 COMSOL simula ion The CFD-simula ions a e pe o med in COMSOL Mul iphysics 4.2a using a p e iously desc ibed model o non-New onian luid in ape ed mic ochannels o which he heo e ical backg ound is desc ibed elsewhe e. 2 The unde lying mic ochannel geome y is adjus ed o he exis ing de ices (250 µm channel wid h, 5:1 ape ing-, 2:1 leng h a io, 100 µm heigh ) a a low a e o 2000 µl h-1. The sc eening mainly in ol ed a a ia ion o he hological p ope ies o he sample which a e desc ibed by he Cole-Cole- (o C oss-) equa ion: 29-31 147 wi h he ze o shea iscosi y , he high-shea iscosi y , he in e nal elaxa ion ime and he powe law exponen cha ac e izing he shea hinning be ween and . The modeling pa ame e s a e lis ed in he ollowing Tab.1. Table 1 Lis o modeling pa ame e s o he CFD-simula ions including he heological pa ame e a ia ions. Pa ame e Value Flow a e , µl h-1 2000 Flow speed , m s-1 0.0222 Ze o shea iscosi y , Pa s 19522; 195.22; 1.9522 High shea iscosi y , Pa s 0.1 In e nal elaxa ion ime , s 388.5; 38.85; 3.885 Powe law exponen 0.99 Densi y , kg m-3 998.2 Tempe a u e , K 293.15 Channel wid h , µm 250 Channel heigh , µm 100 Tape ing a io 10:1 Tape ing leng h a io 2:1 De ice ope a ion a he PETRAIII synch o on (P03 beamline, HASYLAB/DESY). The mic o luidic SAXS scanning expe imen s a e pe o med a he Mic o- and Nano ocus X- ay Sca e ing beamline (MiNaXS) in Hambu g, Ge many. The luid low is con olled using high-p ecision sy inge pumps (Nemesys, Ce oni GmbH). A e he collec ion o he ele an backg ound da a om he mic ochannel and an ex ended low equilib a ion ime a a low a e o 200 µl h-1, he SAXS-mapping scans a e commenced. The mic o ocused X- ay beam o hese scans has a wa eleng h o λ = 0.095 nm wi h a wid h o 20 µm and a heigh o 30 µm. The SAXS pa e ns a e eco ded a a dis ance o 4840 mm wi h in eg a ion ime o 0.5 s using a digi al de ec o (Pila us by Dec is) wi h a pixel size o 172 µm by 172 µm wi h. 148 Conclusions In conclusion, we ha e pe o med a undamen al s udy o he e-o ien a ion o aniso opic pa icles in con ined geome ies o a wide ange o expe imen al condi ions. We also demons a ed in g ea de ail how mul iple analysis echniques can be used and combined in a complemen a y way aking ad an age o he highly ep oducible design con ol p o ided by mic o luidics. Nex o he he e-demons a ed co ela ion o luid dynamics o s uc u al in o ma ion, he combina ion o complemen a y me hods is also applicable and ex endable o o he me hods and mo e expe imen s, such as low ield analysis o mixing expe imen s, ollowing syn he ic eac ion kine ics and mapping nuclea ion and g ow h p ocesses. 5,6,9,32-34 Using complemen a y analysis ools, including small-angle X- ay sca e ing, mic opa icle image elocime y, pola iza ion mic oscopy and CFD-simula ions, we could iden i y design ules o p edic ing and con olling he pe pendicula e-o ien a ion o aniso opic pa icles a e na ow sec ions. A ela i e anking acco ding o he e ec i eness o pe pendicula alignmen o aniso opic pa icles is de i ed as: ape ing a io (y-o ien ing ex ensional a e) > concen a ion (plug- low and non-New onian beha io ) > leng h a io (wall shea su ace) > low speed (wall shea in ensi y). The egions o pe pendicula o ien a ion can be con olled wi h u he p ecision by uning he heological p ope ies o he sample, i. e. like he ze o shea iscosi y and he in e nal elaxa ion ime. Consequen ly, he pe pendicula o ien a ion o aniso opic pa icles can be maximized by inc easing he ape ing a io, sho ening he ape ing, minimizing he low speed and inc easing he sample’s plug- low by inc easing i s high ze o shea iscosi y and sho ening i s in e nal elaxa ion ime o ice e sa o main aining and inc easing pa allel alignmen . In summa y, he he e p esen ed s udy p o ides condi ions o con olling he o ien a ion o aniso opic pa icles. A s ong pa allel o ien a ion is o g ea impo ance o he spinning o ibe s because i s ypically leads o inc eased c ys allini y in he ibe and, hence, be e mechanical pe o mance. 12,35 Howe e , he pe pendicula o ien a ion o pa icles can also be o g ea in e es o applica ions ha in ol e elec ical o he mal anspo pe pendicula o a su ace, such as in isola ing ma e ials o hyb id sola cells. 36-38 149 Re e ences 1. D. C. Du y, J. C. McDonald, O. J. A. Schuelle , and G. M. Whi esides, Anal. Chem., 1998, 70, 4974–4984. 2. M. T ebbin, D. S einhause , J. Pe lich, A. Bu e , S. V. Ro h, W. Zimme mann, J. Thiele, and S. Fö s e , PNAS, 2013, 110, 6706–6711. 3. A. O en, S. Kös e , B. S u h, A. Snigi e , and T. P ohl, J Synch o on Rad, 2005, 12, 745–750. 4. R. Ba e , M. Faucon, J. Lopez, G. C is obal, F. Des emau , A. Dodge, P. Guillo , P. La al, C. Masselon, and J.-B. Salmon, Lab Chip, 2006, 6, 494–499. 5. T. P ohl, A. O en, S. Kös e , R. Doo z, B. S u h, and H. M. E ans, Biomac omolecules, 2007, 8, 2167–2172. 6. H. M. E ans, R. Doo z, S. Koes e , B. S u h, and T. P ohl, Bull Pol Acad Sci-Te, 2007, 55, 217–227. 7. R. Doo z, H. E ans, S. Kös e , and T. P ohl, Small, 2007, 3, 96–100. 8. K. N. To , B. Ves e gaa d, S. S. Nielsen, D. Snakenbo g, M. G. Jeppesen, J. K. Jacobsen, L. A le h, and J. P. Ku e , Anal. Chem., 2008, 80, 3648–3654. 9. M. E. B ennich, J.-F. Nol ing, C. Dammann, B. Nöding, S. Bauch, H. He mann, T. P ohl, and S. Kös e , Lab Chip, 2011, 11, 708. 10. S. M. Tahe i, S. Fische , M. T ebbin, S. Wi h, J. H. Sch öde , J. Pe lich, S. V. Ro h, and S. Fö s e , So Ma e , 2012, 8, 12124. 11. A. Bu e , A. Ro hki ch, R. Döh mann, V. Kö s gens, M. M. Abul Kashem, J. Pe lich, G. He zog, M. Schwa zkop , R. Geh ke, P. Mülle -Buschbaum, and S. V. Ro h, J. Synch o on Rad (2012). 19, 647-653 [doi:10.1107/S0909049512016895], 2012, 1–7. 12. T. Vad, J. Wul ho s , T.-T. Pan, W. S einmann, S. Dab inghaus, M. Becke s, G. Seide, T. G ies, W. F. C. Sage , M. Heidelmann, and T. E. Wei ich, Mac omolecules, 2013, 130705124618000. 13. H. H. Yang and S. R. Allen, Ad anced Fibe Spinning Technology, Woodhead Publica ions, Abing on, 2000. 14. J. A. Cuculo, J. F. Ho e , Q. Zhou, and Q, S uc u e Fo ma ion in Polyme ic Fibe s, Hanse Ga dne Publica ions, 1s edn. 2004. 15. N. F. Bouxsein, L. S. Hi s , Y. Li, and C. R. Sa inya, Applied Physics …, 2004. 16. K. J. Humph y, P. M. Kulka ni, D. A. Wei z, J. F. Mo is, and H. A. S one, Phys. Fluids, 2010, 22, 081703. 17. C. F. Ca lbo g, T. Ha aldsson, K. Öbe g, M. Malkoch, and W. an de Wijngaa , Lab Chip, 2010, 11, 3136. 18. P. Wägli, A. Homsy, and N. F. de Rooij, Senso s and Ac ua o s B: Chemical, 2011, 156, 994–1001. 19. E. Sollie , C. Mu ay, P. Maoddi, and D. Di Ca lo, Lab Chip, 2011, 11, 3752. 20. G. M. Whi esides and Y. Xia, Angew. Chem. In . Ed., 1998, 37, 550–575. 21. Y. Xia and G. M. Whi esides, Annual e iew o ma e ials science, 1998, 28, 153–184. 22. S. Fö s e , L. Apos ol, and W. B as, J Appl C ys , 2010, 43, 639–646. 150 23. S. Fö s e , S. Fische , K. Zielske, C. Schellbach, M. Sz ucki, P. Lindne , and J. Pe lich, Ad ances in Colloid and In e ace Science, 2011, 163, 53–83. 24. JPIV 1.0: Open sou ce so wa e package o pa icle image elocime y. A ailable a www.jpi . ennemann-online.de. Accessed July 1, 2013. 25. S. Fö s e and T. Plan enbe g, Angew. Chem. In . Ed., 2002, 41. 26. M. An onie i and S. Fö s e , Ad anced Ma e ials, 2003, 15, 1323–1333. 27. S. Fö s e , M. Kon ad, and P. Lindne , Phys. Re . Le ., 2005, 94. 28. C. Hanske, J. E a h, C. Küh , M. T ebbin, C. Schneide , A. Wi emann, and A. Fe y, Zei sch i u Physikalische Chemie, 2012, 226, 569–584. 29. K. S. Cole and R. H. Cole, J. Chem. Phys., 1941, 9, 341. 30. K. S. Cole, J. Chem. Phys., 1942, 10, 98. 31. M. M. C oss, J Coll Sci Imp U Tok, 1965, 20, 417–437. 32. L. Pollack, M. Ta e, N. Da n on, J. Knigh , S. G une , W. Ea on, and R. Aus in, PNAS, 1999, 96, 10115–10117. 33. L. Pollack, M. W. Ta e, A. C. Finne ock, C. Kalidas, S. T o e , N. C. Da n on, L. Lu io, R. H. Aus in, C. A. Ba , S. M. G une , and S. Moch ie, Phys. Re . Le ., 2001, 86, 4962–4965. 34. R. Russell, I. S. Mille , M. W. Ta e, L. W. Kwok, B. Naka ani, S. M. G une , S. Moch ie, V. Pande, S. Doniach, D. He schlag, and L. Pollack, P oceedings o he Na ional Academy o Sciences, 2002, 99, 4266–4271. 35. M. E. Kinahan, E. Filippidi, S. Kös e , X. Hu, H. M. E ans, T. P ohl, D. L. Kaplan, and J. Wong, Biomac omolecules, 2011, 12, 1504–1511. 36. G. Kaune, M. A. Rude e , E. Me walli, W. Wang, S. Coue , K. Schlage, R. Röhlsbe ge , S. V. Ro h, and P. Mülle -Buschbaum, ACS Appl. Ma e . In e aces, 2009, 1, 353–360. 37. M. Rawolle, M. A. Niede meie , G. Kaune, J. Pe lich, P. Lellig, M. Memesa, Y.-J. Cheng, J. S. Gu mann, and P. Mülle -Buschbaum, Chem. Soc. Re ., 2012, 41, 5131. 38. M. A. Rude e , S. Guo, R. Meie , H.-Y. Chiang, V. Kö s gens, J. Wiede sich, J. Pe lich, S. V. Ro h, and P. Mülle -Buschbaum, Ad . Func . Ma e ., 2011, 21, 3382–3391. 151 152 Supplemen al In o ma ion Figu e S1 Expe imen al se up o cap u ing s uc u al in o ma ion o complex luids in si u a he mic o ocus beamline P03/MiNaXS a he highly b illian synch o on sou ce PETRA III (HASYLAB, DESY). 11 The ypical expe imen al se up consis o h ee undamen al elemen s: mic o ocused X- ay beam, X- ay compa ible mic o luidic de ice and X- ay de ec o . This combina ion o mic o luidics and X- ay mic obeams is a powe ul expe imen al echnique which o e s a ious ad an ages compa ed o non-con inuous SAXS expe imen s. Figu e S2 Pola iza ion mic oscopic image o he illus a ion ha he pe pendicula micelle o ien a ion e ec is main ained a e he addi ion o small polys y ene ace sphe es (1.65 µm adius, PDI 1.01). 153 7. T ebbin, M.; Thiele, J.; Wi h, S.; Benecke, G.; Bu e , A.; Abul Kashem, M.; Pe lich, J.; Mülle -Buschbaum, P.; Ro h, S. V.; Fö s e , S. HASYLAB Use 's Mee ing 2012 - Sa elli e Mee ing: S a us and Pe spec i es o Small Angle X- ay Sca e ing a DESY, DESY Hambu g, Janua y 25.-27., 2012. "S a us o he mic o luidic T-SAXS P ojec a P03/MiNaXS". 8. T ebbin, M.; Thiele, J.; Wi h, S.; Bu e , A.; Pe lich, J.; Benecke, G.; Mülle - Buschbaum, P.; Ro h, S. V.; Fö s e , S. 25 h ECIS Con e ence (Eu opean Colloid and In e ace Socie y) & 45 h Biennial Mee ing o he Ge man Colloid Socie y, Technical Uni e si y o Be lin, Sep embe 4.-9., 2011. "Pa icle-o ien a ion con ol in mic o luidic de ices". 9. T ebbin, M.; Bu e , A.; Pe lich, J.; Benecke, G.; Thiele, J.; Wi h, S.; Kö s gens, V.; Rawolle, M.; He zog, G.; Mülle -Buschbaum, P.; Ro h, S. V.; Fö s e , S. HASYLAB Use 's Mee ing 2011 - Sa elli e Mee ing: S a us and Pe spec i es o Small Angle X- ay Sca e ing a DESY, DESY Hambu g, Janua y 26.-28., 2011. "La es esul s om mic o luidics a MiNaXS". 10. T ebbin, M.; Fische , S.; Tahe i, S.; Meye , A.; Wi h, S.; Thiele, J.; Fö s e , S. 2nd TUM-HASYLAB Colloquium “The me al-polyme in e ace”, DESY Hambu g, No embe 2.-3., 2010. "A mic o luidic sample en i onmen a a mic o ocus beamline - basics and pe spec i es". Pos e p esen a ions: 11. T ebbin, M.; K üge , K.; DePon e, D.; Ro h, S. V.; Schulz, J.; Chapman, H. N.; Fö s e , S. XFEL Use 's Mee ing 2013, DESY Hambu g, Janua y 23.-25., 2013. "Mic o luidic liquid je sys ems". 12. Wi h, S.; Fü s , C.; T ebbin, M.; Chen, X.; Ba z, C.; Ro h, S. V.; Fö s e , S. HASYLAB Use 's Mee ing 2013, DESY Hambu g, Janua y 23.-25., 2013. "Scanning s uc u al e olu ion o lyo opic phases wi h mic o luidics & mic o ocus SAXS". 13. T ebbin, M.; Ho mann, E.; Blüm, C.; Heideb ech , A.; Lang, G.; Alb ech , G.; Kü ne , M.; F ey ag, A.-S.; Ba gel, H.; Scheibel, T.; Fö s e , S. 4 h Scien i ic Semina o he No h-Ba a ia Bioma e ials Alliance (NBBA), Uni e si y o E langen, No embe 27., 2012. "Spide silk ibe o ma ion in mic o luidic de ices". 256 14. T ebbin, M.; Wi h, S.; S einhause , D.; Pe lich, J.; Ro h, S. V.; Zimme mann, W.; Thiele, J.; Fö s e , S. Mic o luidics 2012, EMBL Heidelbe g, July 25.-27., 2012. "Mic o luidics mee s mic o ocus SAXS: as sc eening and co ela ion o complex luid beha io wi h s uc u al in o ma ion". 15. T ebbin, M.; Thiele, J.; Wi h, S.; Benecke, G.; S einhause , D.; Koe s gens, V.; Bu e , A.; Kashem, M. A.; Pe lich, J.; Mülle -Buschbaum, P.; Ro h, S. V.; Fö s e , S. HASYLAB Use 's Mee ing 2012, DESY Hambu g, Janua y 25.-27., 2012. "Mic o luidics mee s mic o ocus SAXS: pa icle-o ien a ion con ol s udied in- si u". 16. T ebbin, M.; Thiele, J.; S einhause , D.; Pe lich, J.; Ro h, S. V.; Fö s e , S. 1s Bonn Humbold Awa d Winne s' Fo um o he Alexande on Humbold Founda ion - “F on ie s in Mac omolecula and Ma e ial Science”, Bonn, Oc obe 12.-16., 2011. "Mic o luidics mee s mic o ocus SAXS: pa icle-o ien a ion con ol s udied in- si u". 17. T ebbin, M.; Thiele, J.; Wi h, S.; Pe lich, J.; Ro h, S. V.; Fö s e , S. Bay eu h Polyme Symposium ‘11, Uni e si y o Bay eu h, Sep embe 11.-13., 2011. "Pa icle-o ien a ion con ol in mic o luidic de ices". 18. Thiele, J.; T ebbin, M.; Wi h, S.; Pe lich, J.; Fö s e , S. Mak omolekula es Kolloquium, Uni e si y o F eibu g, Feb ua y 24.-26., 2011. “Shea o ien a ion in mic o luidic channels”. 19. Thiele, J. ; T ebbin, M.; Fö s e , S. 44 h Biennial Mee ing o he Ge man Colloid Socie y, Uni e si y o Hambu g, Sep embe 28.-30, 2009. "P epa a ion o monodispe se block copolyme esicles ia low ocusing in mic o luidics". 257 258 10 Acknowledgemen s I would like o exp ess my deepes g adi ude o my supe iso P o . D . S ephan who ga e me he oppo uni y o wo k on bo h exci ing and challenging esea ch opics whils con inuously suppo ing me in e e y imaginable aspec . S ephan, I hank you om my hea . I would also like o hank all my iends, colleagues and cowo ke s a he Uni e si y o Bay eu h, he Uni e si y o Hambu g and a DESY. I would like o hank all he people who con ibu ed o his wo k, especially all my s uden s who wo ked wi h me on many esea ch p ojec s and who con ibu ed g ea ly o he esea ch in he Fö s e g oup. Las ly, I would like o hank my amily and iends o hei ongoing suppo and wisdom. I can’ ell how much I app ecia e wha you ha e done o me. 259 260 11 (Eidess a liche) Ve siche ungen und E klä ungen (§ 5 N . 4 P omO) Hie mi e klä e ich, dass keine Ta sachen o liegen, die mich nach den gese zlichen Bes immungen übe die Füh ung akademische G ade zu Füh ung eines Dok o g ades unwü dig e scheinen lassen. (§ 8 S. 2 N . 5 P omO) Hie mi e klä e ich mich dami ein e s anden, dass die elek onische Fassung meine Disse a ion un e Wah ung meine U hebe ech e und des Da enschu zes eine gesonde en Übe p ü ung hinsich lich de eigens ändigen An e igung de Disse a ion un e zogen we den kann. (§ 8 S. 2 N . 7 P omO) Hie mi e klä e ich eidess a lich, dass ich die Disse a ion selbs ändig e ass und keine ande en als die on mi angegebenen Quellen und Hil smi el benu z habe. Ich habe die Disse a ion nich be ei s zu E langung eines akademischen G ades ande wei ig einge eich und habe auch nich be ei s diese ode eine gleicha ige Dok o p ü ung endgül ig nich bes anden. (§ 8 S. 2 N . 9 P omO) Hie mi e klä e ich, dass ich keine Hil e on gewe bliche P omo ionsbe a e n bzw. - e mi le n in Ansp uch genommen habe und auch kün ig nich nehmen we de. .................................................................................................... O , Da um, Un e sch i 261