biofouling and cleaning of microchannels

Biofouling and cleaning of Microchannels by J.Ponmozhi Dissertation submitted to UNIVERSIDADE DO PORTO For the degree of Doctor in Chemical and Biological Engineering Supervised by: Dr. João Mario Miranda Co-Supervised by: Prof. João M. Campos Prof. Filipe Mergulhão Centro de Estudos de Fenómenos de Transporte Departmento de Engenharia Química Faculdade de Engenharia da Universidade do Porto Portugal January, 2016 ! II! ! III! ! IV! ! V! Abstract Biofouling in microdevices is not as visible as biofouling seen in macroscale scale devices. There is a necessity to monitor and amplify the microscopic events that human eye cannot see, into the visual scale, in order to have an efficient and continuous functioning of the microdevices. The main objective of the present thesis is to investigate the biofouling in microchannels, with emphasis on the initial adhesion step. A cell suspension of Escherichia coli flows through the microchannel and cell adhesion is monitored by visualization along the experiment. The channels used comprise a set of diverse designs, from simple rectangular straight microchannels, to microchannels with multiple constrictions, all belonging to different microfluidic devices. The flow parameters and temperature used for the biofouling study mimic real human body in vivo conditions. The microchannels used in the experiments were produced in polydimethylsiloxane (PDMS) polymer by soft lithography technique using molds manufactured by two methods, namely xurography and photolithography (SU-8 molds). A new in-house technique was developed to enable the insertion of a polymeric surface into one of the walls of the microchannel. The developed technique has a simple protocol to follow, without the need for costly apparatus and also without any constrain in the fabrication process. The adhesion of E. coli over the inserted bottom wall surface was visualized through the microscope and the images captured at different time intervals were processed in order to find the adhesion rate. Experiments on biofouling to understand initial adhesion phenomenon were carried following three guiding vectors: 1- range of wall shear stresses that occur in the human body; 2 - different biomedical polymer surfaces; 3 - microchannel geometries incorporating constrictions as the main hot spot adhesion region. A small region of interest, out of the whole length of the microchannel, was monitored to find the adhesion rate for a particular wall shear stress, surface or geometry. Biofouling on six different polymeric surfaces were studied along with glass (used as reference) surface for two different shear stresses to correlate the adhesion rate with surface properties (hydrophobicity and zeta potential). The adhesion rates on polymeric surfaces on micro and macro scales were compared. When similar wall shear stresses are used in both scales, similar adhesion rates are obtained whatever the surface ! VI! enabling the use of different scales for adhesion tests, depending on expertise and equipment availability. In the adhesion studies on different microchannel geometries, the whole length of the microchannel was analyzed to quantify the adhesion before, in and after the constriction regions. The results obtained were stimulating and they can be applied in different microfluidic devices and application fields. The cleaning of microchannels with the help of antibiotics was also investigated. The conjugate effects of wall shear stress and antibiotics contribute to the removal of adhered cells. Data on dead and live cells removed using antibiotics were obtained. To clarify the stability and predictability of the flow patterns near the observation region, numerical simulations were made with the commercial code ANSYS Fluent CFD package (version 14.5) by solving Navier–Stokes equations. In straight channels, the wall shear stress was found to be constant along the flow direction. With the new in-house technique developed to incorporate the polymer patch in the microchannel, a lowered surface is attained near the region of interest. In this region, the wall shear stress suffers a small transition, which the experimental results demonstrate to have a negligible effect on adhesion. Keywords:* Cell adhesion; Biofouling; Biomedical coatings; Biomicrofabrication; Computational fluid dynamics; Microfluidics; Escherichia. Coli; Antibiotic cleaning; Constrictions; Multiple constrictions channel; Converging channel; Biofouling hot spots! ! VII! Resumo A incrustação biológica em microdispositivos não é tão visível como visto a incrustação biológica em dispositivos escala macro-escala. Há uma necessidade de monitorar e amplificar os eventos microscópicos que o olho humano não pode ver, na escala visual, a fim de ter um funcionamento eficiente e contínua dos microdispositivos. O principal objetivo do presente trabalho é investigar o estágio inicial do processo de formação de biofilmes, a adesão inicial de bactérias a superficies, em microcanais. Durante as experiências, de adesão inicial de bactérias realizadas uma suspensão de células de Escherichia coli flui através do microcanal e a adesão de bactérias é monitorizada através da visualização por microscopia. Foram estudados vários designs de canais, em diferentes dispositivos microfluídicos, desde canais reactangulares simples a canais com múltiplas constrições. Os parâmetros de escoamento e temperatura utilizadas para o estudo foram seleccionados para mimetizar as condições in vivo do corpo humano. Os microcanais utilizados nas experiências foram produzidas em polidimetilsiloxano (PDMS) por litrografia suave a partir de moldes fabricados por dois métodos alternativos, nomeadamente a fotolitografia (moldes de SU-8) e a xurografia. Uma nova técnica de fabricação de microcanais foi desenvolvida para permitir a inserção de uma superfície polimérica numa das paredes do microcanal. Foi desenvolvido um protocolo simples de seguir, sem a necessidade de equipamentos dispendiosos. A adesão de E. coli à superfície inserida foi visualizada através do microscópio e as imagens captadas ao longo do tempo foram processadas para determinar a taxa de adesão em diferentes materiais. As experiências para compreender fenómeno de adesão inicial foram realizados de acordo com três linhas orientadoras: 1- a gama de tensões de corte na parede foi selecionada para ser idêntica à que ocorre no corpo humano; 2 – os materiais selecionados foram consideradas superfícies de polímeros com aplicação biomédica; 3 – as geometrias estudadas têm relevância para microdispositivos de análises clínicas.. Duante as experiências, uma pequena região do microcanal foi monitorizada para determinar a taxa de adesão para cada taxa de corte, geometria ou superfície. A adesão em vidro (referência) e em seis superfícies poliméricas diferentes foi estudada para duas taxas de corte diferentes com o objectivo de tentar correlacionar a taxa de ! VIII! adesão com as propriedades das superfícies (hidrofobicidade e potenciais zeta). As taxas de adesão em superfícies poliméricas à micro e à macro escalas foram comparadas. Quando são usadas tensões de corte na parede semelhantes em ambas as escalas, são obstidas taxas de adesão semelhantes, qualquer que seja a superfície. Este resultado sugere que as experiências de adesão inicial podem ser feitas a qualquer das escalas estudadas, sendo que a escolha do procedimento experimental e da escala utilizada depende da experiência da equipa de investigação e da disponibilidade de equipamento. Nos estudos de adesão em diferentes geometrias, foi determinada a adesão ao longo do microcanal sendo possível identificar o efeito de constrições e variação da secção recta do canal. Os resultados obtidos são estimulantes e eles podem ser aplicados em diferentes microdispositivos e campos de aplicação. A limpeza de microcanais com o auxílio de antibióticos também foi investigada. Os efeitos conjugados de tensão de corte e antibióticos contribuem para a remoção de células aderentes. Para esclarecer a estabilidade e previsibilidade dos padrões de escoamento junto da região de observação, foram feitas simulações numéricas com o código comercial ANSYS Fluent (versão 14.5), o qual resolve as equações de Navier-Stokes. Em canais simples, de secção recta constante, a tensão de corte de corte é constante ao longo da direcção do escoamento. Com a nova técnica in-house desenvolvido para incorporar uma superfície polimérica no microcanal, forma-se uma região rebaixada na vizinhança da região de interesse. Nesta região, a tensão de corte na parede sofre um pequeno decréscimo, que os resultados experimentais demonstram que têm um efeito desprezável sobre a adesão celular. Palavras-chave: Adesão celular; Biosujamento biológico; Biomicrofabricação; Mecânica dos Fluidos Computacional; Microfluidica; Escherichia. Coli; Limpeza por antibiotico; Constrições; Canal com multiplas constrições; Canal convergente; Biosujamento em pontos críticos. ! IX! To My life partner Shanmugam Dhinakaran, and Kids D.Ganesh and D. Shanmugam who always helped me, in every aspect of life. ! XVI! ! XVII! List of Figures Figure!1.1.!!Visible!biofouling!in!day-to-day!life:!(a)!Biofouling!in!kitchen!sink;!(b)! Biofouling!in!teeth.!!..................................................................................................................!1! Figure! 1.2.! ! Figure! 1.1.! Macroscale! biofouling:! (a)! Ship! hull! with! macro! organisms;!(b)!Condenser!tubes!with!microorganisms!!!.........................................!2! Figure!1.3.!!Microscale!biofouling:!(a)!Fouled!and!clogged!microchannel!used!for! cell! encapsulation! (LOC! device);! (b)! Fouled! point-of-care!device! that! can! both!monitor!blood!glucose!levels!and!deliver!medication!that!reduces!high! sugar!levels.!!................................................................................................................................!2! Figure!1.4.!Size!scale! representation! of!the!E.#coli#bacterium! and!microchannel! used!for!fouling!studies!in!microchannels.!....................................................................!3! Figure!1.5.!Advantages!of!microfluidics!related!to!initial!adhesion.!............................!8! Figure!1.6.!Pictorial!presentation!of!the!thesis!outline.!..................................................!14! Figure!2.1.!Scanning!Electron!Microscopic!image!of!a!staphylococcal!biofilm.!....!22! Figure!2.2.!Images!obtained!with!an!inverted!phase!contrast!microscope!of!cells! (300*250!μm)!cultivated!on!PEI!polymeric!films!after!90!h..!.............................!27! Figure! 2.3.! Microstructured! flow! channels,! square! and! offset! patterned,! for! adhesion!studies.!....................................................................................................................!28! Figure! 3.1.! Procedure! followed! to! make! the! mold! for! the! production! of! microchannels.!........................................................................................................................!44! Figure!3.2.!Process!followed!for!making!micromolds!in!various!films:!(a)!Cut!the! microchannel! in! the! vinyl! film! and! peel! off! the! unnecessary! film! area;! (b)! Peel!off!the!microchannel!mold!pattern!with!a!transport!tape;!(c)!Press!the! microchannel!pattern!along!with!the!transport!tape!in!a!Petri!dish;!(d)!Peel! off!the!transport!tape!leaving!the!micromold!in!the!Petri!dish.!!.......................!45! Figure! 3.3.! Micromolds! created! with! Xurographic! technique! using:! (a)! a! blue! film;!(b)!a!red!film.!.................................................................................................................!46! Figure! 3.4.! Rectangular! PDMS! microchannel! developed! from! Xurography! mold! used!for!initial!adhesion!study.!........................................................................................!46! Figure!3.5.!PDMS!device!fabrication!procedure:!(a)!cross!section!of!SU-8!mold!/! any!mold!with!a!positive!relief;!(b)!mixture!of!PDMS!and!curing!agent!in!a! ratio!of!5:1!is!poured!over!the!mold!and!kept!in!the!oven;!(c)!cured!PDMS!is! peeled!off!from!the!mold!and!access!ports!are!created!with!the!syringe!tips;! (d)!PDMS!layer!containing!the!channel!structure!is!bonded!to!the!glass!slide! covered!with!the!thin!layer!of!PDMS!and!placed!in!the!oven!for!12!hours!to! seal!the!channels.!...................................................................................................................!47! Figure! 3.6.! The! different! polymer! surfaces! captured! through! the! microscopic! camera.!!......................................................................................................................................!49! Figure!3.7.!(a).!Plain!Glass!slide!or!any!substrate;!(b).!Polymer!coated!over!the! substrate;!(c).!Scotch!tape!pasted!over!polymer!and!PDMS!coated!over!the! tape;!(d).!The!scotch!tape!removed!and!the!PDMS!with!channel!obtained!by! soft!lithography!is!bonded.!................................................................................................!50! Figure!3.8.!Experimental!set!up!for!fouling!study.!............................................................!51! Figure!3.9.!a)!Schematic!illustration!of!the!static!contact!angle!using!sessile!drop! method,! captive! bubble! method! and! Wilhelmy! method;! b)! Surface! characterization!based!on!hydrophobicity!and!hydrophilicity.!........................!52! ! XVIII! Figure! 3.10.! Microchannel! representation! and! mesh! details:! a)! Microchannel,! showing! the! lowered! surface! in! grey,! the! region! of! interest! in! black! and! domain! limits;! b)! Profile! representing! the! level! of! the! upper! and! lower! surfaces!of! the!channel;!c)!3D!representation! of!the! numerical!domain;!d)! Lowered!surface!detail;!e)!Microchannel!cross-section!outside!the!lowered! region;!f)!Cross-section!available!to!the!flow!in!the!lowered!surface!region. !........................................................................................................................................................!56! Figure!3.11.!Wall!shear!stress!(WSS)!in!the!lowered!surface!region!for!a!nominal! wall!shear!stress!of!1!Pa.!.....................................................................................................!58! Figure!3.12.!Velocity! magnitude!in!the! midplane! in!the!lowered! surface!region! for!a!nominal!wall!shear!stress!of!1!Pa.!........................................................................!58! Figure!3.13.!Wall!shear!stress!along!the!centreline!at!the!bottom!surface!in!the! lowered! region! for! a! nominal! wall! shear! stress! of! 1! Pa.! Figure! show! numerical!predictions!(symbols),!predicted!based!on!analytical!equation!6! and! predictions! with! corrections! based! on! equation! 7.! Local! domain! coordinates!are!used!to!represent!the!distance!from!the!domain!inlet.!........!59! Figure!3.14.!!Raw!and!processed!images,!size!312!×!233!μm2,!for!counting!cells !.......................................................................................................................................................!.61! Figure!3.15.!Zoomed!view!of!the!images!of!size!92.38!×!73.41!μm2!from!Figure! 3.14.!..............................................................................................................................................!62! Figure!3.16.!!Zoomed!view!of!the!processed!PDMS!image!of!size!92.38!×!73.41! μm2,!including!cells!(3),!PDMS!surface!disturbance!(1)!and!noise!(2).!..........!63! Figure!3.17.!E.#coli!adhesion!on!PLLA,!PS!and!PDMS!over!time.!..................................!64! Figure! 3.18.!Cell! density! after! 1800! s! of! adhesion! for! three! different! shear! stresses! (a)! 0.02! Pa,! (b! )! 0.2! Pa! and! (c)! 1! Pa.!The! square! represents! the! lowered!region!in!the!microchannel.!.............................................................................!65! Figure! 4.1.! ! Flow! chart! describing! the! selection! of! polymers! for! the! initial! adhesion!study.!.......................................................................................................................!76! Figure!4.2.!Straight!microchannel!with!dimensions.!........................................................!78! Figure! 4.3.! Wall! shear! stress! in! the! microchannel! obtained! by! numerical! simulation.!.................................................................................................................................!79! Figure!4.4.!The!microchannel!with!the!polymeric!surface!and!an!enlarged!view!of! the!focusing!area.!...................................................................................................................!80! Figure!4.5!(a).!Adhesion!of!E.coli!at!different!time!points!on!different!surfaces!of! size!201.83!×!191.05!μm2!...................................................................................................!81! Figure! 4.6! (b).! Adhesion! experiments! (three! trials)! made! during! 30! mins! in.! a! cellulose!acetate!coated!microchannel.!........................................................................!83! Figure! 4.7.! Average! adhesion! of! E.# coli! on! cellulose! acetate! along! time;! the! adhesion!rate!is!578.55!nº!of!cells/(cm2.s).!................................................................!84! Figure! 4.8.! Fouling!rate! in! a! PDMS! surface! versus! wall! shear! stress! applied.! Ranges! of! wall! shear! stress! in! human! circulatory! settings:! 1)! vena! cava,! aorta,!veins!urinary!catheter,!bladder!implant;!2)!venules,!infra-renal!aorta! [43]!;!3)!!arteries;!4)!!capillaries.!.....................................................................................!86! Figure!4.9.!E.#coli!bacteria!adhesion!along!experiment!for!a!low!wall!shear!stress! (0.0208!Pa).!Linear!trend!during!1200!s.!....................................................................!87! Figure! 4.10.! E.# coli! bacteria! adhesion! along! experiment! for! a! high! wall! shear! stress!(2!Pa).!The!adhesion!stops!after!200!s.!...........................................................!87! Figure!4.11.!Trendline!fitted!up!to!200!s!for!the!plot!in!Figure!4.10.!........................!87! ! XIX! Figure!4.12.!E.#coli!bacteria!fouling!rates!on!different!polymers!at!two!wall!shear! stresses!(0.02!and!0.01!Pa)!versus!contact!angle.!...................................................!88! Figure!4.13.!E.#coli!bacteria!fouling!rate!versus!hydrophobicity!of!the!surface….89! Figure!4.14.!E.#coli!bacteria!fouling!rate!versus!zeta!potential………………………….90! Figure! 4.15.! E.# coli! bacteria! adhesion! rates! on! PA,! glass,! PDMS,! CA! and! PLLA! obtained! in! the! microchannel! (black! bars)! and! in! the! PPFC! (white! bars).! Error! bars! shown! for! each! surface! represent! the! standard! deviation! from! three!independent!experiments![33].!...........................................................................!92! Figure!5.1.!Sudden!constriction!channel!design!(not!to!scale).!.................................!104! Figure!5.2.!Converging!channel!design!(not!to!scale).!...................................................!105! Figure!5.3.!Multiple!sudden!constriction!channel!design!(not!to!scale)..!..............!105! Figure! 5.4.! ! 2D! representation! of! the! numerical! domains:! (a)! constriction! geometry;!(b)!multiple!constrictions;!(c)!converging!microchannel.!...........!108! Figure! 5.5.! Schematic! representation! of! monitoring! the! whole! length! of! the! microchannel!with!the!microscope!objective!for!every!500!μm!interval.!...!109! Figure!5.6.!Adhesion!along!the!2000!μm!constriction!microchannel!at!different! shear!stresses.!.......................................................................................................................!110! Figure!5.7.!Microscopic!image!of!2000!μm!constriction!at!different!spots!along!its! length! for! 0.2! Pa! shear! stress:! (a)! before! the! constriction! (uniform! adhesion);! (b)! start! of! the! constriction! (sudden! contraction! -! sudden! increased!adhesion);!(c)!end!of!the!constriction!(decreased!adhesion);!(d)! after!the!constriction!(sudden!expansion!–!highly!decreased!adhesion);!(e)! location! of! a,b,c! and! d! adhesion! zones! along! the! 2000! μm! constriction! microchannel.!........................................................................................................................!111! Figure!5.8.!Average!adhesion!along!the!length!of!the!microchannel!with!500!μm! constriction!for!0.2!Pa.!......................................................................................................!!113! Figure!5.9.!Average!adhesion.!along!the!length!of!the!microchannel!with!2000!μm! constriction!for!0.2!Pa!........................................................................................................!113! Figure!5.10.!Average!adhesion!along!the!length!of!the!microchannel!with!5000! μm!constriction!for!0.2!Pa!................................................................................................!113! Figure!5.11.!Microscopic!images:!!(a)!no!adhesion!zone;!and!(b)!lower!adhesion! zone!(depleted!region)!for!5000!μm!constriction!microchannel!length!at!0.2! Pa;!(c)!location!of!a!and!b!adhesion!zones.!...............................................................!114! Figure! 5.12.! Adhesion! along! the! length! of! the! microchannel! with! multiple! constrictions:!(a)!τ#w=!0.2!Pa;!(b)!τw#=!1!Pa.!..............................................................!115! Figure!5.13.!Adhesion!in!a!multiple!constriction!channel!for!0.2!Pa.!......................!116! Figure!5.14.!Adhesion!in!a!multiple!constriction!channel!for!1!Pa.!.........................!116! Figure!5.15.!Converging!microchannel!for!0.2! Pa:!(a)!Adhesion! with!respective! channel! width! at! each! location! along! the! length;! (c)! Adhesion! with! wall! shear!stress!at!each!location!along!the!length!.........................................................!117! Figure!5.16.!Converging!microchannel!for!1!Pa:!(a).!Adhesion!along!the!length!of! the! converging! microchannel! for! 1! Pa! at! the! inlet,! (b)! Adhesion! with! respective! channel! width! at! each! location! along! the! length,! (c)! Adhesion! with!wall!shear!stress!at!each!location!along!the!length.!...................................!118! Figure! 5.17.! Microscopic! image! for! the! converging! microchannel! at! 0.2! Pa:! (a)! End!of!the! converging! region;! (b)!Beginning!of!the!expansion! region! after! the! convergent;! (c)! Location! of! a! and! b! on! zones! along! the! converging! microchannel!.........................................................................................................................!119! ! XX! Figure!5.18.!Microscopic!image!for!the!converging!microchannel!at!1!Pa:!(a)!End! of! the! converging! region;! (b)! Beginning! of! the! expansion! region! after! the! convergent).! (c)! Location! of! a! and! b! zones! along! the! converging! microchannel!.........................................................................................................................!120! Figure!5.19.!!Magnitude!of!the!velocity!in!a!500!µm!long!constriction.!The!flow! rate!is!3.42!×!10-11!m3s-1.!The!constriction!width!is!10!µm!and!the!channel! depth!is!100!µm.!...................................................................................................................!121! Figure! 5.20.! ! Wall! shear! stress! in! a! 500! µm! long! constriction.! The! flow! rate! is! 3.42!×!10-11!m3s-1.!The!constriction!width!is!10!µm!and!the!channel!depth!is! 100!µm.! The!nominal! wall!shear!stress!in!the!upstream!channel!is!0.2!Pa.! The!wall!shear!stress!calculated!by!the!numerical!method!in!the!upstream! channel!is!0.31!Pa.!The!wall!shear!stress!in!the!constriction!is!19.2!Pa.!......!121! Figure!5.21.!!Magnitude!of!the!velocity!in!a!converging!channel.!The!flow!rate!is! 8.58!×!10-11.!m3s-1.!................................................................................................................!122! Figure!5.22.!!Wall!shear!stress!in!the!converging!channel.!The!flow!rate!is!8.58!×! 10-11.!m3s-1.!..............................................................................................................................!122! Figure! 5.23.! ! Velocity! magnitude! in! a! section! of! a! multiple! constrictions! microchannel.!The!flow!rate!is!3.42!×!10-11!m3s-1.!The!constriction!width!is! 10!µm!and!the!channel!depth!is!100!µm.!max!is!13.8!..........................................!123! Figure! 5.24.! ! Velocity! magnitude! in! a! section! of! a! multiple! constrictions! microchannel.!The!flow!rate!is!3.42!×!10-11!m3s-1.!The!constriction!width!is! 10!µm!and!the!channel!depth!is!100!µm.!...................................................................!123! Figure! 5.25.! Fouling! rates! at! different! locations! as! mentioned! in! 1,2,3,4! plots! (location!of!A!to!I!points)!with!their!local!wall!shear!stress.!1-2000!µm;!2!-! 5000!µm!constriction!for!a!nominal!shear!stress!of!0.2!Pa;!3!and!4!-!multiple! constritions!.............................................................................................................................!124! Figure!6.1.!Schematic!representation!of!the!Y!channel!used!for!determining!the! fouling!and!cleaning!rates.!...............................................................................................!134! Figure!6.2!!Visualization!of!the!fouled!microchannel!with!E.#coli!bacteria!after!2.5! hours!from!the!beginning!of!fouling.!...........................................................................!137! Figure!6.3.!Microscopic!images!of!the!Y!–shaped!microchannel!shown!in!Figure! 6.1! at! different! times! when! ciprofloxacin! antibiotic! was! used,! with! corresponding!cell!count!by!Image!J!software.!.......................................................!138! Figure!6.4.!Continuation! of! Microscopic! images!of!the!Y! –shaped! microchannel! shown! in! Figure! 6.1! at! different! times! when! ciprofloxacin! antibiotic! was! used,!with!corresponding!cell!count!by!Image!J!software.!.................................!139! Figure! 6.5.! Plot! showing! cell! adhesion! and! cleaning! with! citrate! buffer! alone.! Experiments!were!conducted!at!a!wall!shear!stress!of!0.03!Pa.!......................!140! Figure!6.6.!Plot!showing!the!number!of!adhered!cells!per!unit!area.!!Results!are! an!average!of!three!independent!experiments.!Standard!deviation!between! the!individual!assays!is!indicated!with!error!bars.!................................................!141! Figure!6.7.!Plot!showing!percentage!of!cell!removal!in!microchannel!and!number! of! cells! killed! (by! CFU! plate! count! method)! at! the! end! of! 2.5! hours! in! microchannel.!........................................................................................................................!142! ! XXI! List of Tables Table!3.1.!Polymers!and!solvents!used!for!preparing!polymeric!solutions!...........!48! Table!3.2.!Hydrodynamic!conditions!.......................................................................................!55! Table!3.3.!Surface!characterization!with!Hydrophobicity!and!Zeta!potential!.......!60! Table!4.1!Polymers!application!.................................................................................................!77! Table!4.2.!!Wall!shear!stress!in!different!locations!of!the!human!body!and!also!in! biomedical!devices!................................................................................................................!84! Table!4.3.!Wall!shear!stresses!and!correspondent!adhesion!rates!for!PDMS! microchannels!.........................................................................................................................!86! Table!4.4.!Hydrophobicity!for!different!surfaces!along!with!adhesion!rates!for! 0.01!Pa!and!0.02!Pa.!..............................................................................................................!89! Table!4.5!Zeta!potential!values!for!each!polymeric!surface!..........................................!90! Table!5.1.!Dimensions!of!the!channels!with!a!sudden!contraction!..........................!104! Table!5.2!Shear!stress!at!the!region!of!interest!and!flow!rates!values!employed!in! studies!with!different!microchannels!.........................................................................!107! Table!5.3!Adhesion!rates!and!respective!shear!stresses!(from!chapter!4)!..........!107! Table!5.4.!Wall!shear!stresses!in!a!microchannel!with!a!500!µm!long!constriction.! The!constriction!width!is!10!µm!and!the!channel!depth!is!100!µm.!..............!122! Table!5.5.!Adhesion!rates!with!local!and!nominal!wall!shear!stress!for!the!points! mentioned!in!Figure!5.25.!................................................................................................!125! ! XXII! ! XXIII! Nomenclature Greek Surface contact angle of water Surface contact angle for Bromonafthalene Surface contact angle for formamide ρ Density τ Shear stress τls Shear stress of lowered surface τw Shear stress at the wall µ Viscosity of the fluid Acronyms PDMS Polydimethylsiloxane PA Polyamide PLLA Poly-l-lactide acid PS Polystyrene PEO Polyethylene oxide PEI Polyethyleneimine POC Point of care MEMS Microeletromechanical systems LOC Lab-on-a-chip DCM Dichloromethane Roman Re Reynolds number x, y, z Cartesian coordinates Q flow rate θ w θα −B ( ) ( ) F θ ! XXIV! Le Entrance length Lc Constriction length Lin Length of the inlet section Lout Length of outlet section H Height of the microchannel Hls Height of the lowered surface Lw Length of the wide section of the multiple constriction section. Lconv Length of the converging channel Wc Width of the constriction W Width of the microchannel ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! XXV! ! CHAPTER(1(-(INTRODUCTION(((((! ! ( ! !J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!6! b. Does fouling depend on the scale on which it operates viz. micro scale and macro scales? c. If the deposits are thick, will it be easier, or more difficult, to clean them? So far, many types of measures were developed to combat biofilms, like using antibiotics to kill or using modified surfaces to inhibit initial adhesion of cells. The process of biofilm formation starts with initial adhesion [30]. If there is clear understanding of the initial adhesion mechanism, biomaterials that are less colonized by bacteria can be developed by material science engineers and researchers [31]. So there arises a need to quantify and understand the initial adhesion phenomenon over different surfaces with respect to time. In the last few years there has been a large improvement in the optical techniques to measure velocity profiles, record the initial adhesion and biofilm formation by monitoring them online. So, optical techniques can be relied on as capable tools to study the fouling phenomena. 1.3.2 Advantages of microfluidics ! Microfluidics is a versatile multidisciplinary field that is attractive for a wide range of applications [32]. Pertaining to the advantages of microfluidics made use in initial adhesion studies, they are remarkable. The study made with microfluidics for the initial adhesion is almost the same as the study made for biofilm development where the biofilm formation platform is continuously monitored to know the response of C. albicans and P. pastoris biofilms to different shear stress conditions [33]. It is possible to have a real time monitoring in a hydrodynamic environment in high throughput devices in chemical and biological engineering fields. It is also possible to mimic the human in vivo condition. With the automation of the microfluidic devices, it is possible to have good repeatability and reproducibility with increased selectivity and sensitivity in the experiments when compared to their macro counterparts. CHAPTER(1(-(INTRODUCTION(((((! ! ( ! !J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!7! The microfluidic devices can be manufactured with fine details applying developed microfabrication techniques and in an inexpensive way applying xurography microfabrication method. The microfabrication techniques that developed in 1980’s has driven MicroElectroMechanical Systems (MEMS) that has more commercial values in mechanical, chemical, biomedical domains [34-36], supports other domains such as Automotive (active suspension, vehicle security systems, automatic door locks), Space exploration field (pressure sensors, micro-power sources and turbines), etc. Xurography introduced by Bartholomeusz et al. [37] can produce the micromolds, bypassing the expensive photolithography technique for areas where precision in microchannels construction is not required. With the help of the microfabrication techniques, the microchannel width with respect to the bacterium characteristic dimension can be easily manipulated. Very small volumes of reagents are enough, since surface effects are dominating over volume effects in the biofouling study. Due to the compact size of the microchannel, the inertial forces become negligible and the viscous force is prevalent resulting in low Reynolds numbers. At same hydrodynamic conditions, microchannels encounters low Reynolds number while turbulent flows are often encountered in macro flow systems. The real time monitoring is made with transparent microchannel; the monitoring was made at the same location for a period of time and also scanned through out the length of the channel. A range of shear stress can be used in a microfluidic-closed system, where outside disturbances and infections could be avoided making it beneficial for initial adhesion study. The temperature of the very small flow volumes assays can be maintained at the in vivo body temperature. The advantages of microfluidics pertaining to the initial biofouling study implemented in this thesis are listed in Figure 1.5. The microfluidic approach developed in this thesis has the capacity further elucidate the so far unknown mechanism of adhesion, which would surely subsides fouling related problems in many fields. CHAPTER(1(-(INTRODUCTION(((((! ! ( ! !J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!8! Figure 1.5. Advantages if microfluidics related to initial adhesion 1.4 Objectives ! The biofouling study is made in a microscale basically for creating new research tools for determining the efficiency of developing and developed microscale biomedical devices. The main objectives of this thesis are summarised as follows: a. Develop a microfluidic set up that could mimic the biomedical scenario based on in vivo temperature and wall shear stress; ! b. Analyse,! establish! and! finalize! different! polymers,! with! biomedical! applications,!to!be!used!as!microchannel!walls,!based!on!their!durability,! spincoatability!and!ease!of!fabrication;! CHAPTER(1(-(INTRODUCTION(((((! ! ( ! !J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!9! c. Develop fabrication methods to produce microchannels with polymeric walls, adequate for adhesion research in polymers with biomedical applications; d. Perform an initial adhesion study, to understand the biofouling phenomenon that happens in vivo implant surfaces or catheters, lab-on-chips (LOC) and point of care diagnostics (POC) microfluidic systems; e. In all the microscale equipment (POC, LOC, etc.) constrictions are the hot spot zones where blockage occurs due to biofouling. Different geometries will be developed, compared, analysed and the adhesion results will be identified taking in account their relevance in different biomedical applications; f. Analyse the difference between micro and macro scales adhesion with different polymeric surfaces that are commonly employed in biomedical devices. Compare the difference in the results obtained between micro and macro scales platforms to determine whether the dimensions of the real systems do not need to be mimicked and the experiments can be performed either at the micro as at the macro scales; g. Investigate the cleaning phenomenon of fouled microchannels by antibiotics as an aid the shear stress cleaning. 1.5. Methods ! The biofouling study was made in PDMS microchannels with very small volume of E. coli suspension, which was prepared with overnight cell cultures by conventional microbiology methods. The polymeric surfaces were prepared in-house with wellestablished methods, available in the literature, and they were characterized through hydrophobicity studies, before and after the biofouling experiment, to test the durability of their properties. A Zeta potential study was also made to characterize the polymer and the E. coli cell property surfaces. The biofouling study was made varying different parameters: wall shear stress, surface properties, geometries with and without constrictions, along with comparison of biofouling in micro and macro scales. The cleaning of biofouled microchannels was done with two antibiotics to understand its killing and removal properties. CHAPTER(1(-(INTRODUCTION(((((! ! ( ! !J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!10! The images of the biofouling process during the experiments (30 minutes) were used for counting adhered E. coli on the bottom wall of the microchannel. Through out the thesis work, all the initial adhesion studies were carried out at human body temperature of 37 °C and for a time period of 30 mins. The experimental methods, necessary for the biofouling study, microfabrication and microscopy methods, were available in our research group, Centro de Estudos de Fenómenos de Transporte (CEFT). Microbiology methods for preparing the cell culture and surface characterization methods were available in Laboratório de Processos de Engenharia, Ambiente, Biotecnologia e Energia (LEPABE) group. With the facilities available (CEFT), xurography technique was used to prepare microchannels for certain experiments. A new in-house method was developed for obtaining a polymeric wall in the microchannel. The SU-8 molds were purchased. 1.5.1 Microfabrication methods ! Microchannels were produced in CEFT laboratory with the help of soft lithography technique. SU-8 molds, produced by photolithography process were purchased and used for producing microchannels. Molds were prepared by xurography, which is a novel and rapid technique recently introduced for producing microchannels in different films. A normal cutter plotter was used for making negative features of a microchannel of 100 µm in films that could be easily transported on to a Petridish with a transport sheet, as the cutter plotter can cut films with thickness ranging from 25 to 1000 µm [37]. The films used for the molds are normal adhesive vinyl films that are used in sign industry for cutting graphics [38]. Major reasons for choosing xurography technique are: • Easy to produce microchannels with molds prepared from adhesive vinyl films of different geometries in a very short time; • The whole mold production can be recreated if the adhesion experiments need design changes. • It does not have any time consuming process using chemicals for mold production as photolithographic process; CHAPTER(1(-(INTRODUCTION(((((! ! ( ! !J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!11! • The production cost is inexpressible, making it extremely fit for any basic and infant research operation study. The few disadvantages that can be overlooked for basic research are the resolution, it is not as high as in the standard lithographic techniques, and only the materials cut by the cutter blade can be used for making microchannels. A new in-house microfabrication technique to insert a polymeric wall in the microchannel was developed in the present thesis. ! 1.5.2 Microbiology methods ! E. coli was prepared in LEPABE group following conventional microbiology methods for culture preparation. This strain was used because it had already demonstrated a good biofilm formation capacity [39]. 1.5.3 Microscopy ! The adhesion of E. coli was monitored using a fluorescence inverted microscope, along 30 minutes. The images were captured with a CCD camera in consecutive time intervals. The captured images were used for counting the adhered cells. All the microchannels produced were transparent, one of the criteria in developing microchannels by different techniques in the laboratory. The images captured were stored in tiff format by Leica application suite software. The images stored were processed and the number of E. coli cells was counted with Image J software, a versatile tool for quantifying the fouling. 1.5.4 Characterization of materials and cells ! The prepared polymeric surfaces were characterized for hydrophobicity, durability (adherence to the channel wall throughout the biofouling study) and zeta potential, with the facilities available in LEPABE laboratory. The hydrophobicity was determined by the contact angle between the surface and polar and apolar liquid drops. CHAPTER(1(-(INTRODUCTION(((((! ! ( ! !J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!12! The contact angles were determined automatically by the sessile drop method in a contact angle meter. The surface charges of the polymers and of the E. coli were characterized through the zeta potential. 1.5.5 Computational Fluid Dynamics (CFD) ! The flow in the microchannel was simulated by numerical methods to clarify the stability and predictability of the flow patterns near the observation region. The main drawback while developing the new in-house technique for insertion of the polymer wall in the microchannel was a lowered surface level of approximately 10 µm. This region was analysed with numerical flow simulation to understand whether this lowered surface affects the WSS and velocity fields. Numerical simulations were made with the commercial code Ansys Fluent CFD package (version 14.5) by solving Navier-stokes equations. 1.6. Summary of the thesis ! The thesis is summarised as follows with a brief description along with the pictorial representation Figure 1.6. The present chapter (Chapter 1) is the introduction and section that provides the overall information about the thesis work and structure. Chapter 2 is the literature review section, where the literature on the topics relevant to the thesis is reviewed. This chapter is focused on the influence of surface properties, wall shear stress and geometrical configuration on initial adhesion, with particular emphasis on works developed in microfluidic devices. Chapter 3 is devoted to describe the microchannel fabrication methods. The approaches used to produce the microchannels to study initial adhesion are described. Two types of methods were used to fabricate microchannels in this thesis: soft lithography from SU-8 molds and by combining xurographically-produced molds and CHAPTER(1(-(INTRODUCTION(((((! ! ( ! !J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!13! soft lithography. A method was also developed to add a polymer patch to the wall of the channel to be used in adhesion tests of different polymeric surfaces. Chapter 4 relates the study performed about initial adhesion on different polymeric surfaces. A relation between zeta potential, hydrophobicity and adhesion rate is discussed for different polymeric surfaces. The adhesion rate for a range of wall shear stress, based on applications, is studied for polydimethyl siloxane (PDMS) transparent polymer. Chapter 5 deals with initial adhesion of cells in different locations of microchannels. Here, initial adhesion is quantified on different types of geometries, trying to understand the effect of geometries on cell adhesion, in particular the effect of wall shear stress changes along the channel. The aim of the study is to find the critical points prone to fouling and clogging in microchannel networks as it is mostly relevant to applications such as Micro Electro-Mechanical Systems (MEMS), Lab on Chip Devices (LOC) and Point of Care Systems (POC). Chapter 6 is devoted to study one of the cleaning strategies based on wall shear stress and antibiotics. Ampicillin and ciprofloxacin were prepared in LEPABE group. The adhesion process and detachment process were observed for five hours with a Y shaped xurographically-fabricated microchannel. The detaching influence of two antibiotics over the E. coli that had adhered at a particular wall shear stress was compared. The wall shear stress is kept constant for both fouling and cleaning mechanisms. Unlike other studies, here fouling and cleaning were observed both during two and a half hours. CHAPTER(1(-(INTRODUCTION(((((! ! ( ! !J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!14! Figure 1.6. Pictorial presentation of the thesis outline. CHAPTER(1(-(INTRODUCTION(((((! ! ( ! !J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!15! The principal results of the thesis are summarised in the conclusion chapter (Chapter 7). The future prospective works, regarding this biofouling work, are discussed in the future work chapter (Chapter 8). References ! [1]!J.!Rayner,!R.!Veeh,!J.!Flood,!Prevalence!of!microbial!biofilms!on!selected!fresh! produce!and!household!surfaces,!Int!J!Food!Microbiol,!95!(2004)!29-39.! ! [2]! v.L.! A.,! Microscopical! observations! about! animals! in! the! scurf! of! the! teeth.,! Philos!Trans!R!Soc!Lond!B!Biol!Sci,!14!(1684)!568-574.! ! [3]! S.E.! Coetser,! T.E.! Cloete,! Biofouling! and! biocorrosion! in! industrial! water! systems,!Critical!reviews!in!microbiology,!31!(2005)!213-232.! ! [4]! T.R.! Bott,! Industrial! Biofouling:! Occurrence! and! Control,! Elsevier! Science,! 2011.! ! [5]! M.E.! Callow,! J.E.! Callow,! Marine! biofouling:! a! sticky! problem,! Biologist! (London),!49!(2002)!10-14.! ! [6]! Z.! Song,! L.! Borgwardt,! N.! Hoiby,! H.! Wu,! T.S.! Sorensen,! A.! Borgwardt,! Prosthesis! infections! after! orthopedic! joint! replacement:! the! possible! role! of! bacterial!biofilms,!Orthop!Rev!(Pavia),!5!(2013)!65-71.! ! [7]!P.!Stoodley,!G.D.!Ehrlich,!P.P.!Sedghizadeh,!L.!Hall-Stoodley,!M.E.!Baratz,!D.T.! Altman,!N.G.!Sotereanos,!J.W.!Costerton,!P.!Demeo,!Orthopaedic!biofilm!infections,! Curr!Orthop!Pract,!22!(2011)!558-563.! ! [8]!L.!McLaughlin-Borlace,!F.!Stapleton,!M.!Matheson,!J.K.!Dart,!Bacterial!biofilm! on! contact! lenses! and! lens! storage! cases! in! wearers! with! microbial! keratitis,! J! Appl!Microbiol,!84!(1998)!827-838.! CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 22! exopolymeric matrix mainly composed of polysaccharide material [3]. Biofilm formation starts with cell adhesion to a pre-conditioned surface. The development of biofilms in medical devices is a common problem, which can lead to hospitalization, revision surgery or mortality. Recent reviews provide a source of evidence to the undesirable biofilm formation in medical devices [4-10] as shown in Figure 2.1. These biofilms pose a challenge to the health care community. Currently, there is a spurring development of smart polymers that are used for coating biomedical implants, artificial organs, lab on chip surfaces, implantable drug delivery systems [11], in order to reduce biofouling. They will give way for the development of a next generation of biomedical devices, which are less prone to fouling. Figure 2.1. Scanning Electron Microscopic image of a staphylococcal biofilm [5] Initial adhesion is a largely unexplored research area, since full-formed biofilms have been attracting most of the researchers. The majority of the initial adhesion studies so far are based on different surfaces; different bacterial strains and different environmental conditions and they were performed using different experimental procedures. Only a handful amount of research was done with the help of microfluidic CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 23! devices. But keeping in mind the advantages of microfluidics [12] and the need for point-of-care devices (POC), it is expected that more research on biofilms will be carried out on these platforms. There is plenty of scope relating adhesion studies to microfluidic devices, since, in a near future, the likely trend is that every individual will be using POC and lab-on-a-chip (LOC) tools for safety and security reasons. Cell micro-processing devices have been developed to deal with biological fluids containing cells. Flow cytometry [13] allows for the continuous measurement and discrimination of individual cells flowing across a sensor. Moreover, microdevices have been developed for blood fractionation [14], cell trapping and analysis [15, 16], removal of pathogens from blood [17] and transformation of bacteria using plasmids [18]. 2.2 Historical review on adhesion in microchannels Adhesion of microorganisms exists everywhere, from teeth to ship hulls. This phenomenon just needs a surface and water to materialize. The earliest studies about adhesion of microorganisms were performed at the macroscale, where the event can be visualized by naked eye. Research on macrofouling has received significant funding [19-22], as it was realised that fouling control was very expensive in different scenarios. Research on biofilm formation in smaller (micro) scales has started to develop [23, 24], but studies on initial adhesion in microchannels are scarce. Although initial adhesion and detachment can be studied under a microscope by different means and methods [25] and despite the advantages of microfluidics, the research on initial adhesion in microchannels is still in its infancy. Earlier in 1995, platelet aggregation in patients with Behcet's disease was studied using silicon microchannels [26]. In order to reduce adhesion, PVC was tested with strains of Pseudomonas aeruginosa by Triandafillu et al. [27]. These authors concluded that oxygen plasma treated surfaces yields a hydrophilic surface and reduces the number of adhering bacteria up to 70%. A study on attachment and detachment with four types of living cells namely T47D, U937, CaCo2 and NCTC 2544 was made by Zhang et al. [28] who concluded that applying an optimal range of CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 24! wall shear stresses can enhance adhesion. In a microfluidic study, it was concluded that adhesion depends on the shape of the adhering cell/particle [29]. Tousi et al. [30] reported that adhesion does not depend on cell deformation, cell signalling nor heterogeneous distribution of adhesion molecules, but primarily depends on the hemodynamic forces. They made both in vivo and in vitro studies (with a PDMS based microvascular device) to confirm their findings. In 2012, a numerical study of the lateral migration and deformation of leukocytes in a microchannel was investigated, to open up the possibility of deploying microchannels for deformabilitybased flow cytometry studies [31]. Recently, in 2013, a study was made to analyse the random motion of mammalian cells in a continuous microchannel, in order to aid in the design of scaffolds for tissue engineering (Young-Gwang et al. [32]). Additionally, Hojin Ha and Sang-Joon Lee [33] studied platelet aggregation and stenosis in a microchannel, whereas Schnegas et al. [34] investigated, by 3D modelling and computational fluid dynamics simulations (CFD), the force required to detach a eukaryotic cell from a microchannel. Other adhesion studies in microfluidic devices, without any analysis of the shear flow field, were made by several researchers [35, 36], which are of less importance to this thesis. 2.3 How do cells adhere? In a channel, the cells start to flow freely, but depending on the shear stresses field, they progressively transit from free flow to dragging flow or rolling flow to reversible adhesion and then irreversible adhesion. This concept is explained and supported by most of the published studies [37]. However, a recent work by Wang et al. [38] claims that, when the cell size is lower than 1 µm, adhesion is not influenced by the hydrodynamic flow. In the present thesis, the bacteria used are the Gram negative Escherichia coli. The characteristic dimensions of this bacteria range between 1 and 3 µm. In most of the studies addressing biofouling, researchers aimed at developing strategies to inhibit biofilm growth. An alternative strategy consists on delaying initial adhesion so that the onset of a biofilm can be retarded. For this purpose, a lot more information must CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 25! be gathered regarding the first phases of biofilm formation and for that purpose it is of paramount importance to monitor the adhesion process from the beginning. 2.4 Current techniques used to observe and quantify initial adhesion The cell adhesion process was studied by different methodologies [39] since the 1960’s. These techniques include: • Internal reflection techniques [40]; • Total internal reflection microscopy (TIRM) which was initially developed to observe single particles [41]; • Total internal reflection fluorescence microscopy (TIRFM) [42]; • Infrared spectroscopy techniques [43]; • Quartz crystal microbalance analysis [44]; • Centrifugation techniques [45] including measurements based on gravity [46] and other spinning disc techniques [47]; • Cytodetacher technique [48] to measure the detachment forces; • Atomic force microscopy (AFM) [49]. All of these techniques have their relative advantages and disadvantages but most of them are sophisticated, require a complex experimental set up, and must be applied without any disturbance during the experimental execution. 2.5 Initial adhesion studies based on applied wall shear stresses Different cell adhesion studies have been performed based on applied shear stress. An interesting study was made by Wesley et al. [50] who captured cells (using a low flowrate of 1µL/min) in a microchannel. They tried to quantify two types of cells, namely, HL-60 and U-937 at a given wall shear stress. A review study was made by Bianchi et al. [51] about leukocyte adhesion on a microchannel surface based on applied wall shear stresses!featuring several high-throughput solutions for adhesion CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 26! monitoring. A PDMS multi shear microchannel design was used to quantify the cell adhesion on various protein-coated surfaces, varying the applied wall shear stress [52]. Different wall shear stresses in the same microchannel were achieved by varying the width along the length of the microchannel. A large range of wall shear stresses, from 0 to 200 Pa, was achieved with this multi-shear microchannel device. Another study to isolate tumour cells using microsystems was performed with two types of cells (BT-20 and MDA-MB-231) at different flowrates up to 3 ml/min [53]. It was reported that the detachment rate of BT-20 was higher and the attachment rate lower when compared to MDA-MB-231 cells at a constant flowrate of 0.3 ml/min. Further, the cell dynamics of cancer cells in free motion, rolling adhesion and firm adhesion were studied in microchannels [37]. Cheung et al. [54] used epithelial-cell-adhesion- molecule (EpCAM) functionalized microchannels to separate two types of cancer cells, namely breast cancer cells (MDA-MB-231) and prostate cancer cells (PC3N). They concluded that: 1- higher hydrodynamic loading decreased cell adhesion along the channel length; 2 - the distance travelled by the cells before the adhesion is linearly proportional to the rolling velocity; and 3 - the capture efficiency of the cancer cells is smaller for high wall shear stresses. Cell rolling behaviour was monitored in real time for wall shear stresses ranging from 0.05 to 0.3 Pa, in a microsystem that mimics the endothelial microvasculature of a bone marrow [55]. The adhesion and detachment processes of different mutant strains of P. aeruginosa [matrix (pelA), type I (cupA 1), type IV pili (pilC) and flagellum defective (flgK)] were influenced by the intensity of the applied wall shear stresses. The pilC mutation negatively affected bacterial adhesion when wall shear stress was increased [56]. Other works have reported different parameters affecting cellular adhesion such as culture time and surface geometry [57, 58]. 2.6 Effects of the surface material on the initial adhesion. The review on bacterial biofilms by Hall-Stoodley!et! al. [59, 60] gives a better understanding about biofilms in different surface materials and the need to prevent them. The physicochemical properties of the polymer surface have a strong influence in cell attachment and detachment. Research based on different polymer coatings, CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 27! with three medically relevant bacteria, namely, P. aeruginosa, Staphylococcus epidermidis, and Staphylococcus aureus, was made. The results show that cell adhesion can decrease up to 50% according to the type of polymer coating used [61]. In a study in microchannels using three different surfaces, namely SiO2 uncoated surface and octadecyltrimethoxysilane (OTMS) or N-(triethoxysilylpropyl)-O- polyethylene oxide urethane (TESP) coated on SiO2, the microchannel with uncoated surface was clogged rapidly due to cell adhesion [62]. The OTMS coated surface had a 75% decrease in the flowrate whereas the TESP coated surface experienced a 20% decrease. Although there are numerous publications addressing cell adhesion in polymeric surfaces [63-65], as shown in Figure 2.2. where adhesion of rat neuronal cells are tested with polyethyleneimine (PEI) biocompatible polymer. The research on adhesion in polymer surfaces in microchannels is scarce. The reason may be due to the expensive microfabrication techniques and problems associated with cell visualization in these systems. Figure 2.2. Images obtained with an inverted phase contrast microscope of cells (300*250 µm) cultivated on PEI polymeric films after 90 h. [65] 2.7 Effects of the geometry on the initial adhesion. The effect of the geometry of the microchannel on the initial adhesion was the least studied topic. The impact of channel geometry is important as outlined in a patent by Manalis et al. [66] using a microchannel to capture diseased! blood cells for high! throughput!diagnosis. Chang et al. [50] attempted to capture particular type of cells (HL-60 and U-937) with the help of different types of microstructures, like pillars arranged in different manners in a microchannel. Two types of pillar structures were CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 28! designed to quantify the cells, as shown in Figure 2.1. They reported that the offset pillar structures were able to capture more cells in the first minutes than the square pillar structures. From the third minute on, the capture efficiency was more or less the same for both types of geometries. Liu et al. [57], performed another interesting work with four different microchambers with identical surface area, placed in circle, in a polygon with 8 sides, 12 sides and 16 sides. This study was made to demonstrate that the geometry is responsible for adhesion changes. It was observed that fouling was more effective in complex geometries than in simple geometries. Cox et al. [62] made a cell adhesion test section using hexagonal microstructures to create restricted flows in the microchannels. They studied the flow rate changes over time due to adhesion. Yan et al. [67] fabricated U-shaped microsieves inside a microfluidic chip for the adhesion of tumor cell-targeted microbubbles and breast cancer cells at a wall shear stress of 0.05 Pa. They concluded that accumulation of targeted microbubbles was strongly influenced by flow velocity and that higher retention of microbubbles on cell surfaces was attained at flow velocities lower than 0.03 cm/s. Figure 2.1. Microstructured flow channels, square and offset patterned, for adhesion studies [50]. ! CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 29! 2.8 Cleaning studies based on antibiotics Cleaning or removal of cells is essential for chemical and biological engineering applications [68-71]. Until now, there are vague cleaning methods of the unnecessary cells or particles adhered onto a surface. Due to this scenario, the most advantageous lab-on-chip devices cannot be appropriately used by most of the companies. Industrially, the most commonly addressed and understood problem has been the fabrication cost, which was in part solved by Whiteside’s work [72-74]. However, PDMS fabrication explained by Whiteside would not reduce the cost when it comes to mass production as PDMS fabrication methods do not scale-up. Another problem has been the fouling in lab-on-chip devices and this remains to be solved [75]. A very recent paper by Schoenitz et al. [76] concluded that particulate flows could be controlled in micro devices, resulting in intensified processes involving solid suspensions. In the present thesis, the cleaning of microchannels using antibiotics is explored. ! 2.9 Conclusions The literature on adhesion of cells in micro devices was reviewed taking into account the main topics: wall shear stress effects, surface properties and geometries at the microscale. There are few studies about initial adhesion at the microscale and so there is plenty of scope for these studies. This review gives also relevant information about recent insights. Accordingly, the initial adhesion on microfluidic platforms depends on a variety of chemical and physical parameters, which differ from surface to surface, from geometry to geometry and from flow to flow. Due to the complexity of the phenomena, any adhesion should be monitored in real time to give more insights according to its own specificity. In vivo conditions, where the surface hydrophobicity of the cell changes to a greater extent due to different grown conditions, should be mimicked in microfluidic platforms to get accuracy in the adhesion quantification. The future work is to create microfluidic platforms through which the real environment can be mimicked and the phenomena (adhesion, CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 30! cell removal, and cell quantification) monitored in real time. The present thesis tries to contribute to this challenge. Two important and critical points are the effect of the geometry at the microscale, because different shapes and constrictions are prone to different adhesions, and also to effective cleaning methods. The present thesis focuses in particular on these two themes. References [1] M. Saadatian-Elahi, R. Teyssou, P. Vanhems, Staphylococcus aureus, the major pathogen in orthopaedic and cardiac surgical site infections: a literature review, Int J Surg, 6 (2008) 238-245. [2] W.F. McCoy, J.D. Bryers, J. Robbins, J.W. Costerton, Observations of fouling biofilm formation, Can J Microbiol, 27 (1981) 910-917. [3] R.M. Donlan, Biofilms: Microbial Life on Surfaces, Emerg Infect Dis, 8 (2002) 881-890. [4] J.W. Costerton, P.S. Stewart, E.P. Greenberg, Bacterial Biofilms: A Common Cause of Persistent Infections, Science, 284 (1999) 1318-1322. [5] L.G. Harris, R.G. Richards, Staphylococci and implant surfaces: a review, Injury, 37 Suppl 2 (2006) S3-14. [6] K.F. Kong, C. Vuong, M. Otto, Staphylococcus quorum sensing in biofilm formation and infection, Int J Med Microbiol, 296 (2006) 133-139. [7] C.R. Arciola, D. Campoccia, P. Speziale, L. Montanaro, J.W. Costerton, Biofilm formation in Staphylococcus implant infections. A review of molecular mechanisms and implications for biofilm-resistant materials, Biomaterials, 33 (2012) 5967-5982. CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 31! [8] T. Bjarnsholt, M. Alhede, M. Alhede, S.R. Eickhardt-Sorensen, C. Moser, M. Kuhl, P.O. Jensen, N. Hoiby, The in vivo biofilm, Trends Microbiol, 21 (2013) 466- 474. [9] A.K. Seth, M.R. Geringer, S.J. Hong, K.P. Leung, T.A. Mustoe, R.D. Galiano, In vivo modeling of biofilm-infected wounds: a review, J Surg Res, 178 (2012) 330-338. [10] S. Esposito, S.M. Purrello, E. Bonnet, A. Novelli, F. Tripodi, R. Pascale, S. Unal, G. Milkovich, Central venous catheter-related biofilm infections: An up-to-date focus on meticillin-resistant Staphylococcus aureus, Journal of Global Antimicrobial Resistance, 1 (2013) 71-78. [11] J.C. Middleton, A.J. Tipton, Synthetic biodegradable polymers as orthopedic devices, Biomaterials, 21 (2000) 2335-2346. [12] D. Li, Encyclopedia of Microfluidics and Nanofluidics, Springer, 2008. [13] D. Huh, W. Gu, Y. Kamotani, J.B. Grotberg, S. Takayama, Microfluidics for flow cytometric analysis of cells and particles, Physiological measurement, 26 (2005) R73-98. [14] E. Sollier, M. Cubizolles, Y. Fouillet, J.-L. Achard, Fast and continuous plasma extraction from whole human blood based on expanding cell-free layer devices, Biomed Microdevices, 12 (2010) 485-497. [15] W.-H. Tan, S. Takeuchi, A trap-and-release integrated microfluidic system for dynamic microarray applications, Proceedings of the National Academy of Sciences, 104 (2007) 1146-1151. [16] Q. Zhang, L. Zhu, H. Feng, S. Ang, F.S. Chau, W.-T. Liu, Microbial detection in microfluidic devices through dual staining of quantum dots-labeled immunoassay and RNA hybridization, Analytica Chimica Acta, 556 (2006) 171-177. CHAPTER(2(–(STATE(OF(THE(ART(((((! ! ( ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 38! [69] T.A. Crowley, V. Pizziconi, Isolation of plasma from whole blood using planar microfilters for lab-on-a-chip applications, Lab Chip, 5 (2005) 922-929. [70] L. Nyholm, Electrochemical techniques for lab-on-a-chip applications, Analyst, 130 (2005) 599-605. [71] P.K. Sorger, Microfluidics closes in on point-of-care assays, Nat Biotechnol, 26 (2008) 1345-1346. [72] J.C. McDonald, D.C. Duffy, J.R. Anderson, D.T. Chiu, H.K. Wu, O.J.A. Schueller, G.M. Whitesides, Fabrication of microfluidic systems in poly(dimethylsiloxane), Electrophoresis, 21 (2000) 27-40. [73] J.C. McDonald, G.M. Whitesides, Poly (dimethylsiloxane) as a material for fabricating microfluidic devices, Accounts of chemical research, 35 (2002) 491-499. [74] S.K. Sia, G.M. Whitesides, Microfluidic devices fabricated in poly(dimethylsiloxane) for biological studies, Electrophoresis, 24 (2003) 3563-3576. [75] M. R., When microfluidic devices go bad. How does fouling occur in microfluidic devices, and what can be done about it?, Anal Chem., 77 (2005) 429A- 432A. [76] M. Schoenitz, W. Augustin, S. Scholl, Challenges in cleaning microstructured devices, Food and Bioproducts Processing, 93 (2015) 283-288. ! ! ! ! ! CHAPTER 3 Microchannel fabrication Abstract ! Polymer coated microchannels were fabricated using materials that are commonly used in biomedical implants. In order to mimic in vivo conditions during initial bacterial adhesion, studies were performed at a temperature of 37˚C. This chapter describes the unique in-house fabrication process developed to incorporate the polymer surface in the microchannel by a spin coating method. Using this technique, the implant surface was obtained inside the microchannel and a bacterial suspension of Escherichia coli JM109 (DE3) was pumped into the microchannel to assess the fouling rate of the implant-like surface. Microchannels were fabricated with xurographic technique and the polymers were spin coated over the slide and the microchannels produced so that a leak-tight system was obtained to perform the adhesion studies. The microchannel widths are 450 µm with channel aspect ratio of 4.5. The major drawback of the method is the location of the region of interest in a lowered surface. It was demonstrated by bacterial adhesion experiments and computational fluid dynamics (CFD) simulations that this drawback has a negligible effect on adhesion. It is also shown that shear stress in the region of interest can be calculated by numerical methods and by an analytical equation for rectangular channels. 3.1 Introduction ! Cell adhesion is predominantly determined by the properties of the surface and by the flow characteristics [1]. The present thesis focuses on the study of cell adhesion and cleaning of materials used in biomedical applications, as well as in microdevices in general. Microdevices are considered in two perspectives, as tools for cell adhesion CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 40! research and as objects of research. For these reasons, the construction of the microchannels used to perform the experiments is of primordial importance. Many recent reviews provide a source of evidence to the undesirable biofilm formation in medical devices [2-8]. These biofilms pose a challenge to the health care community. Until now, there are very few bio-adhesion studies to quantify fouling phenomenon on different biomedical surfaces at the micro scale, while there is a spurring development of smart polymers that are used for coating biomedical implants, artificial organs, lab on chip surfaces and implantable drug delivery systems [9], giving way for the development of a next generation of anti-fouling biomedical devices. Ramakrishna et al. [10] discussed different biocompatible polymers used for different types of implants. For example, poly-L-lactide (PLLA) is a biodegradable polymer, which degrades with time without creating harm inside the body [11, 12]. Different types of pin and screws are used for fixing the implants. Autografts namely suspensory fixation, hamstring fixation [13] are used as fixing agents for femoral implants to reduce operation failures. These grafts pins and screws are coated with PLLA. There are several biofilm research studies [14-16] in microchannels, stressing the unique advantages of microfluidics: • The different biological cells can be monitored in real time through visualization techniques; • The assays and different expensive reagents are needed in microliters helping in cost cutting; • The surface can be modified easily and the geometry can be designed according to the application; • The flow remains laminar even at high shear stress values. Rajendrani Mukhopadhyay [17] discussed the sources of fouling in microfluidic devices and advised to keep an eye on fouling at every step of the development process, from the materials to the design of the device. The transmission of a designed pattern on to a substrate through optical means is called photolithography. The desired patterns are transferred to a photoresist liquid CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 41! layer. This layer is coated uniformly over a substrate, exposed and developed with the desired pattern for subsequent processing. The different processes involved in photolithography are expensive and tedious, as they require clean surface preparation, spin coating, alignment for precision output, soft and hard baking, exposure, development, preparing mask alignment with pattern on the wafer’s surface, the photoresist is exposed through the pattern on the mask with high intensity ultraviolet light, developed SU-8 positive microstructures baking [18]. Besides the SU-8 molds prepared by photolithography technique, there are many rapid prototyping techniques (mold making techniques) developed for fabricating microchannels by soft lithography. Duffy et al. [19] implemented a method to produce PDMS channels (soft lithography), which can be used in microscale total analysis systems (µTAS), genetic analysis, clinical diagnostics, drug screening and environmental monitoring. The design, produced using CAD software, is printed in a transparency and used as a mask for making positive relief master molds. The master molds are produced with SU-8 polymer epoxy photoresist described in detail by Blanco et al. (2004) and Che et al. (2002) [20, 21]. The PDMS channels are casted in these molds, which are baked to get well irreversibly sealed channels. The other conventional methods that are commercially used for micromolding are micromilling [22], micropowder blasting [23], hot embossing [24], laser ablation [25] and stereo lithography [26]. Xurography is a technique developed by Bartholomeusz et al. [27], which requires a simple cutter plotter. The microchannels are cut on vinyl films or other type of films according to the application. Positive relief molds of thicknesses between 25 and 1000 µm are generated in the cutter plotter and made ready for casting PDMS microchannels (soft lithographic technique[28] ) in less than 30 minutes. Aspect ratios up to 5.2 for positive features are possible in a 360 µm thick material applying this Xurography technique. The foremost advantage of xurography technique is the reduced capital cost and fast mold fabrication technique. The alternative to xurography is photolithography and the other conventional techniques, both needing expensive machines and expensive clean rooms. Furthermore, in these conventional techniques, any design change in the microchannel geometry requires more than a day to be performed, with long chemical CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 42! procedures and pre and post baking steps. The main disadvantage of xurography is the low resolution, which limits the production of microchannels to dimensions greater than 200 µm. In this thesis, biofouling was studied in rectangular PDMS microchannels to assess the initial adhesion of cells to a surface. Different biomedical polymer surfaces and different configurations of the microfluidic device were tested. The polymers were selected according to their biocompatibility, easy spin coatability on microscopic slides and durability over the experiment. Microchannels were constructed using soft lithographic technique with molds obtained from SU-8 photolithography and xurography techniques. The microchannels produced by xurography are the best for testing the initial adhesion on PDMS surface and also on polymeric surfaces. This technique consumes less time and enables the incorporation of different polymer surfaces in the PDMS microchannels. As already referred, xurography is limited to dimensions higher than 200 µm and so, to study adhesion on smaller micochannels, photolithography was also used to make the molds. 3.1.1 Motivation for fabrication microchannels with polymer coated surfaces ! The motivation to fabricate and study different microchannel polymeric surfaces arises from the literature reviewed concerning failure of different implants due to infection [29] and failure of different implantable sensors [30]. According to this literature, with a better understanding of bacterial adhesion and biofilm formation within the human body may reduce health risks by stopping infections at an early stage, reducing the need for revision surgery, mortality and morbidity of the patients. There are numerous polymeric biomaterials commonly used in biomedical devices and there is a demand to test them under controlled conditions. Microfluidics is an excellent option to test these materials, because it requires small setups, enables controlled operating conditions and the experiments are easily replicated. So far, PDMS based microsystems have received tremendous attention and the review by Jinwen Zhou et al. (2010) focuses on PDMS surface modifications [31]. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 43! In this chapter, a new technique to incorporate a small wall patch into a PDMS microchannel, produced from molds easily made in-house by xurography, is proposed. The wall patch can be made of different polymers. The technique, adapted from a standard PDMS soft-lithography technique, is useful to produce channels for adhesion tests. The spin coating technique is exploited to coat a polymer surface over PDMS. Different polymers were tested taking in consideration the restrictions of the microchannel fabrication methods and also those of the visualization techniques. Only transparent polymers can be used in visualization studies by bright-field microscopy. 3.2 Microfabrication techniques ! Three microfabrication techniques were used in this work: SU-8 photolithography, xurography and soft lithography. Additionally, the soft lithography technique was modified to insert polymeric materials in the microchannels walls. These methods are described in the following sections. ! 3.2.1 SU-8 Photolithography technique ! SU-8 photolithography is a high precision technique to produce microchannels for fouling studies. The fabrication technique is discussed briefly in Sia and Whitesides [32] and in detail in Duffy et al. [19]. The desired microchannel geometries are drawn in AutoCAD drawing software. The AutoCAD drawing is printed on a quartz surface. Hard and durable chrome masks are obtained by sputtering chrome over a quartz substrate [33]. Chrome masks are used for getting microchannel features down to 8 µm as said by Desai et al. [34] and down to 500 nm, which was recently said by Lalanne and Chavel. [35]. The chrome masks are much more expensive than the transparency films [36, 37] (alternative printing substrate to chrome) and they take a much longer time to be fabricated. The molds are produced by a photolithography technique, where light is used to pattern the substrate namely the photocurable epoxy SU-8, a material used for CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 44! producing integrated circuits and microelectronic components [38]. SU-8 is exposed to ultraviolet light through the photomask. The cross-linking process, activated during the exposure, makes SU-8 fully polymerized. The unexposed SU-8, i.e. the unpolymerized SU-8, will get dissolved leaving the coated microchannel design to serve as a positive relief mold for making replicate PDMS microchannels. The rapid prototyping techniques consume much time for producing the positive-relief on the mold surface. The SU-8 photolithography technique enables the production of microchannels with very small dimensions, widths down to 100 µm [39], but the time consumption and requirements of clean environmental room to produce the molds are clearly disadvantages. A pictorial representation of the process of producing molds to prepare microchannels is shown in Figure 3.1. In the present work, the molds were purchased, as the facilities required to produce the SU-8 molds were not available. Figure 3.1. Procedure followed to make the mold for the production of microchannels 3.2.2 Xurography technique ! Bartholomeusz et al. [27], developed a novel rapid prototyping technique for creating micromolds in various thin materials, ranging from 25-1000 µm in thickness, without CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 45! the need for photolithographic or chemical processes. The thin materials are nothing but adhesive films (e. g. vinyl films). The process of making film micromolds is pictorially described in Figure 3.2. After plotting the microchannel pattern over the film using a cutter plotter (GCC Expert 24 vinyl cutter plotter), the unwanted films are removed and the micromold structure is transferred to a Petri dish using a transport tape. This tape can be peeled off leaving the microchannel design in the Petri dish as shown in Figure!3.3. Figure 3.2. Process followed for making micromolds in various films: (a) Cut the microchannel in the vinyl film and peel off the unnecessary film area; (b) Peel off the microchannel mold pattern with a transport tape; (c) Press the microchannel pattern along with the transport tape in a Petri dish; (d) Peel off the transport tape leaving the micromold in the Petri dish. ! Blue and red vinyl films [40] were tested with different cutter speeds (cutter speeds depends on the material and its width) using a GCC Expert 24 vinyl cutter plotter as shown in Figure! 3.3 to produce microchannel micromolds were followed. The microchannel molds were used to make microchannels using a soft lithographic process. A PDMS microchannel casted from the mold is shown in Figure 3.4. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 46! Figure!3.3. Micromolds created with Xurographic technique using: (a) a blue film; (b) a red film. Figure 3.4. Rectangular PDMS microchannel developed from Xurography mold used for initial adhesion study 3.2.3 Soft lithography technique ! The soft lithography technique was used to fabricate microchannels made from polydimethysiloxane (PDMS). The microchannels were fabricated with a homogenous mixture of PDMS (Sylgard 184, Dow Corning). Sylgard 184 is commercially available as a pre polymer kit composed of a PDMS oligomer and a crosslinking agent or curing agent in the ratio of 5:1. A desiccator connected to a vacuum pump was used to remove the air bubbles that formed during the mixing process. The PDMS was poured over the mold and kept in the oven for 20 mins at 80ºC. After curing, the PDMS microchannel was peeled off from the mold. Holes (1mm in diameter) were punched with the help of a syringe tip, through the PDMS replicas at CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 47! both ends of the channel to serve as the flow inlet and outlet. The PDMS microchannels were sealed with PDMS coated glass slide and kept in the oven for approximately 12 hours at 80ºC. The coated PDMS glass was prepared by spin coating, over a glass slide, using a mixture of PDMS and curing agent at a ratio of 20:1. An overview of the fabrication process is depicted in Figure 3.5. This technique allows the construction of mini/micro channels with three walls in PDMS and another wall in PDMS coated on glass. The curing leads to the formation of an irreversible chemical bond between the 5:1 and 20:1 PDMS mixture, ensuring microchannels without leaks. The fabrication process shown in Figure 3.5 is also the general fabrication method for producing microchannels. Figure 3.5. PDMS device fabrication procedure: (a) cross section of SU-8 mold / any mold with a positive relief; (b) mixture of PDMS and curing agent in a ratio of 5:1 is poured over the mold and kept in the oven; (c) cured PDMS is peeled off from the mold and access ports are created with the syringe tips; (d) PDMS layer containing the channel structure is bonded to the glass slide covered with the thin layer of PDMS and placed in the oven for 12 hours to seal the channels. 3.2.4. Insertion of polymer surface in the microchannel ! The transparent and spincoatable polymer solutions were prepared as describe below. The polymer was mixed with the solvent in appropriate mass percentages as indicated in Table 3.1. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 54! the repulsion between similarly charged particles is weak. The zeta potential was calculated using Helmholtz–Smoluchowski equation. ! 3.4. Data analysis with ImageJ ! Adhesion was followed using a fluorescence inverted microscope (DMI 5000M, Leica Microsystems GmbH) with a 40× objective. Microscopic images were captured during 30 mins with a CCD camera (Leica DFC350FX, Leica Microsystems GmbH) with a time interval of 60 s. The image sequence obtained is in tiff format as recorded by Leica Application Suite software. Image processing to count cells over a surface during the experiment was performed using ImageJ, an open source software developed by Wayne Rasband at National Institute of Mental Health (NIH) during 1997. This software allows 8-bit, 16-bit and 32-bit grayscale images and supports a variety of file formats such as, TIFF, GIF, JPEG, BMP, DICOM, FITS and raw data using a URL. It can display a group of images in a single window, image sets are called stacks, and any sequence can be processed in a particular selected window. It uses the standard image processing functions such as contrast, sharpening, smoothing, edge detection and median filtering. !!!All!the!30!images!were!imported!to!Image!J!software![49].!A!low!noise!region! was!selected!in!the!images!using!the!crop!tool.!The!images!were!converted!from! 8! bit! to! 32! bit! to! improve! contrast.! To! set! the! scale! for! processing,! the! pixel! aspect! ratio! was! set! to! one.! Depending! on! the! images,! they! can! be! sharpened! using!mean!filters!with!1!pixel!as!radius.!Then,!the!background!was!subtracted! with! rolling! ball! radius! ranging! from! 1! to! 18! pixels.! A! light! background! was! obtained! with! the! E.# coli!cells! bright! and! visible.! The! brightness! and! contrast! were!fine-tuned!to!get!more!accurate!cell!count.!The!threshold!was!adjusted,!for! the!stack!of!images,!to!generate!a!black!and!white!image,!black!cells!over!a!white! background.!Then!the!cells!were!automatically!counted.!! The cells counted with the help of the ImageJ software were confirmed with manual counting to establish the method developed for the cell counting as appropriate. The cell count result obtained was 99% exact to manual counting. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 55! 3.5. Bacteria and culture conditions ! Escherichia coli JM109 (DE3) was used in the tests, since it has a good adhesion capacity [50]. A starter culture was prepared as described by [51] and incubated overnight. A volume of 60 mL from this culture was centrifuged (for 10 min at 3202 x g) and the cells were washed twice with citrate buffer 0.05 M [52], pH 5. The pellet was then resuspended and diluted in the same buffer to obtain a cell concentration of 7.6 x 107 cells mL-1. 3.6. Flow conditions ! The flow conditions studied are indicated in Table 3.2. With these conditions, the shear stress ranges from 0.02 to 1 Pa, covering the majority of the shear stresses that can be found in the human body [53, 54]. Table 3.2. Hydrodynamic conditions Flow rate(µL/min) Mean Velocity (m/s) Reynolds number Nominal wall shear stress (Pa) 1.35 5.00 x 10-4 0.12 0.02 15 5.56 x 10-3 1.3 0.2 65.1 2.41 x 10-2 5.65 1 ! 3.7. Numerical simulations ! The flow cell was simulated by numerical methods to clarify the stability and the predictability of the flow patterns near the observation region. The microchannel used, represented in Figure 3.10, has a rectangular cross section of 450 x 100 µm and a length of 15 mm. The inlet and outlet have a diameter of 0.44 mm. A section of the microchannel, around the visualization region, was selected for simulation domain (Figure 3.10). This region includes the lowered surface, which results from the fabrication method, where the region of interest is located. The length of the domain is 5 mm. The lowered surface has a length of 3 mm and a width of 3mm. The lowered height is approximately 10 µm. The flow regime was determined by the Reynolds number based on the equivalent diameter of the channel. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 56! ! Re =2 ρ Q (W+H) µ ! (5)! ! where ρ and µ are the density and viscosity of the fluid, respectively, Q the flow rate, W the width of the channel and H the depth of the channel. Figure 3.10. Microchannel representation and mesh details: a) Microchannel, showing the lowered surface in grey, the region of interest in black and domain limits; b) Profile representing the level of the upper and lower surfaces of the channel; c) 3D representation of the numerical domain; d) Lowered surface detail; e) Microchannel crosssection outside the lowered region; f) Cross-section available to the flow in the lowered surface region. Numerical values of the wall shear stress (WSS) were compared with data from the analytical solution for the flow in a parallel plate channel [55]:! τ = µ 3Q 2H 2 ! " #$ % & 2 w ! (6)! !!!Nominal wall shear stress, used to distinguish the experiments, was calculated through the analytical equation 6. The real wall shear stress, as given by the numerical simulation, is slightly different. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 57! The equation for the lowered surface was corrected by the following factor: ! τ ls = τ H Hls ! " #$ % & 2 ! (7)! ! where τ is the wall shear stress in the straight channel (outside the lowered surface) and Hls the depth of the channel in the lowered surface section. Numerical simulations were made with the commercial code ANSYS Fluent CFD package (version 14.5) by solving Navier–Stokes equations. A model of the microchannel was built in Design Modeler 14.5 and was discretized into a grid of 278,000 cells by Meshing 14.5. The QUICK scheme [56] was used for the discretization of the momentum equations and the PRESTO! scheme for the discretization of the pressure terms. The velocity–pressure coupled equations were solved by the PISO algorithm [57]. The no slip boundary condition was considered for all the walls. Simulations were made in steady state mode until convergence. The properties of water (density and viscosity) at 37◦C were used. Corrected numerical results were calculated by applying equation (7). The value of ! used was the wall shear stress of a straight channel (without a lowered surface) previously obtained numerically. ! 3.8. Results ! 3.8.1. Numerical simulation ! Wall shear stress (WSS) at the bottom wall and velocity field at the midplan are represented in Figure 3.11 and Figure 3.12, respectively. These figures show a small velocity decrease in the region of the channel crossing the lowered surface. The velocity at the lateral regions (see Figure 3.10) of the lowered surface is almost zero. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 58! The wall shear stress is also small in the part of the channel that crosses the lowered surface region, where the region of interest is located. Figure 3.11. Wall shear stress (WSS) in the lowered surface region for a nominal wall shear stress of 1 Pa ! ! ! Figure 3.12. Velocity magnitude in the midplane in the lowered surface region for a nominal wall shear stress of 1 Pa. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 59! ! Figure 3.13. Wall shear stress along the centreline at the bottom surface in the lowered region for a nominal wall shear stress of 1 Pa. Figure show numerical predictions (symbols), predicted based on analytical equation 6 and predictions with corrections based on equation 7. Local domain coordinates are used to represent the distance from the domain inlet. Figure 3.13 shows the wall shear stress along the centreline of the bottom wall of the channel. The figure is for the higher nominal WSS studies (1Pa), but results for the other wall shear stresses, not shown here, are similar. Some edge effects are observable, mainly due to the influence of the inlet and outlet boundary conditions. At the inlet, the edge effects are small, revealing that the flow develops in a short length, and enters fully developed in the lowered surface region. In the lowered region, the wall shear stress is smaller. A small transition exists of about 300 µm length. The analytical equation 6 underpredicts the numerical WSS data in the channel outside the lowered region. This result is expected since the analytical equation is exact only for channels with an infinite width. In the present case the ratio between the width and the height of the channel is 4.5, which implies that the velocity is higher than what would be in a channel of infinite width. The correction, equation 7, made on the analytical equation underpredicts the WSS in the lowered surface, while the correction applied to the numerical results predicts it correctly. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 60! ! 3.8.2. Surface characterization ! The surface properties of the different materials fabricated are presented in Table 3.3. Exploring the results, it is observed that except for polyethylene oxide (PEO), the rest of the polymers are hydrophobic. PLLA, PDMS, CA, PA and PS are hydrophobic surfaces (ΔG < 0 mJ.m-2) whereas PEO is hydrophilic (ΔG > 0 mJ.m-2). Additionally, the zeta potential results showed that all the polymers surfaces have a negative charge. The range of surface characteristics assures that the procedure can be used to study a large range of surface parameters. ! ! Table 3.3. Surface characterization with Hydrophobicity and Zeta potential Water θ w ( ) (degree) Bromonafthalene θα −B ( ) (degree) Formamide ( ) F θ (degree) Hydrophobicity ΔG!(mJ!m-2) Zeta potential (mV) PDMS 113.56 ± 0.62 87.61 ± 1.77 111.18 ± 0.61 -61.82 -29.3 PLLA 88.03 ± 1.01 25.58 ± 1.55 68.49 ± 0.95 -65.32 -27.9 PS 80.81 ± 0.82 24.64 ± 0.89 64.32 ± 0.99 -49.56 -29.8 CA 65.24 ± 0.50 22.47 ± 1.05 36.63 ± 2.05 -36.04 -23.4 PA 69.36 ± 0.43 23.63 ± 0.53 48.02 ± 1.24 -37.58 -28.0 PEO 55.54 ± 3.11 34.17 ± 1.28 64.33 ± 0.99 0.35 -11.0 3.8.3. Image processing examples ! Images were processed with ImageJ, as described in section 3.5, and the number of cells attached to the surface was obtained for every minute. Examples of images from adhesion experiments in PDMS, PLLA and PS are shown in Figure 3.14. In the first row it is possible to see images captured at the first minute of the fouling experiment while the second row contains images captured at the end of the fouling experiment (after 30 mins). In the third row are the processed images of those in the second row. Zoomed views of the images presented in Figure 3.14 are in Figure 3.15. A zoomed of the processed image of fouling in PDMS (third row, first column of Figure 3.15 is CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 61! shown in Figure 3.16. It shows bacteria attached and also noise and PDMS surface disturbances. The procedure developed for counting the cells attached eliminates these noise and surface disturbances. First image PDMS First Image PLLA First image PS Last image PDMS Last Image PLLA Last image PS Processed image PDMS Processed Image PLLA Processed image PS Figure 3.14. Raw and processed images, size 312 × 233 µm2, for counting cells ! ! ! ! ! ! CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 62! Figure 3.15. Zoomed view of the images of size 92.38 × 73.41 µm2from Figure 3.14 First image PDMS First image PLLA First image PS Last image PDMS Last image PLLA Last image PS Processed image PDMS Processed image PLLA Processed image PS CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 63! Figure 3.16. Zoomed view of the processed PDMS image of size 92.38 × 73.41 µm2, including cells (3), PDMS surface disturbance (1) and noise (2). 3.8.4. Adhesion results ! The adhesion of E.coli over each polymer surface for 30 minutes was carried out. The data shows that the materials have distinct adhesion behaviour. PS is the material with less adhesion, Figure 3.17, while PDMS has the highest. Additionally, it was observed that bacterial adhesion increases linearly with time. Standard deviations were calculated and plotted as error bars. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 70! [33] V. Perumal, U. Hashim, Chrome mask design for microfluidic fabrication, in: Advanced Materials Research, Trans Tech Publ, 2013, pp. 276-280. [34] T. Desai, M. Ferrari, S.N. Bhatia, BioMEMS and Biomedical Nanotechnology: Volume III: Therapeutic Micro/Nanotechnology, Springer US, 2007. [35] P. Lalanne, P. Chavel, Perspectives for Parallel Optical Interconnects, Springer Berlin Heidelberg, 2013. [36] J. Narasimhan, I. Papautsky, Polymer embossing tools for rapid prototyping of plastic microfluidic devices, J Micromech Microeng, 14 (2004) 96-103. [37] P. Vadgama, Surfaces and Interfaces for Biomaterials, Elsevier Science, 2005. [38] Y. Xia, G.M. Whitesides, Soft Lithography, Angewandte Chemie International Edition, 37 (1998) 550-575. [39] A. del Campo, C. Greiner, SU-8: a photoresist for high-aspect-ratio and 3D submicron lithography, J Micromech Microeng, 17 (2007) R81. [40] K.E. Herold, A. Rasooly, Lab on a Chip Technology: Fabrication and microfluidics, Caister Academic Press, 2009. [41] Z.L. Zhang, C. Crozatier, M. Le Berre, Y. Chen, In situ bio-functionalization and cell adhesion in microfluidic devices, Microelectronic Engineering, 78-79 (2005) 556- 562. [42] J.H. Koschwanez, R.H. Carlson, D.R. Meldrum, Thin PDMS Films Using Long Spin Times or Tert-Butyl Alcohol as a Solvent, Plos One, 4 (2009) e4572. [43] F.J. Mergulhao, G.A. Monteiro, Analysis of factors affecting the periplasmic production of recombinant proteins in Escherichia coli, J Microbiol Biotechnol, 17 (2007) 1236-1241. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 71! [44] K. Merrett, R.M. Cornelius, W.G. McClung, L.D. Unsworth, H. Sheardown, Surface analysis methods for characterizing polymeric biomaterials, J Biomater Sci Polym Ed, 13 (2002) 593-621. [45] Y. Yuan, T.R. Lee, Contact angle and wetting properties, in: Surface science techniques, Springer, 2013, pp. 3-34. [46] B. Janczuk, E. Chibowski, J. Bruque, M. Kerkeb, F.G. Caballero, On the consistency of surface free energy components as calculated from contact angles of different liquids: an application to the cholesterol surface, Journal of colloid and interface science, 159 (1993) 421-428. [47] C.J. Van Oss, Interfacial forces in aqueous media, CRC press, 2006. [48] L.C. Simoes, M. Simoes, M.J. Vieira, Adhesion and biofilm formation on polystyrene by drinking water-isolated bacteria, Antonie Van Leeuwenhoek, 98 (2010) 317-329. [49] C.A. Schneider, W.S. Rasband, K.W. Eliceiri, NIH Image to ImageJ: 25 years of image analysis, Nat Meth, 9 (2012) 671-675. [50] J.M. Moreira, J.D. Araujo, J.M. Miranda, M. Simoes, L.F. Melo, F.J. Mergulhao, The effects of surface properties on Escherichia coli adhesion are modulated by shear stress, Colloids Surf B Biointerfaces, 123 (2014) 1-7. [51] J.S. Teodosio, M. Simoes, L.F. Melo, F.J. Mergulhao, Flow cell hydrodynamics and their effects on E. coli biofilm formation under different nutrient conditions and turbulent flow, Biofouling, 27 (2011) 1-11. [52] M. Simoes, L.C. Simoes, S. Cleto, M.O. Pereira, M.J. Vieira, The effects of a biocide and a surfactant on the detachment of Pseudomonas fluorescens from glass surfaces, Int J Food Microbiol, 121 (2008) 335-341. CHAPTER(3(–(MICROCHANNEL(FABRICATION(((((( ! ! J.!Ponmozhi!–!PhD!!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 72! [53] A.D. Michelson, Preface, in: A.D. Michelson (Ed.) Platelets (Second Edition), Academic Press, Burlington, 2007, pp. xxi. [54] L.S. Ronald, W. University of, Analysis of Pathoadaptive Mutations in Escherichia Coli, University of Washington, 2008. [55] H.J. Busscher, H.C. van der Mei, Microbial adhesion in flow displacement systems, Clin Microbiol Rev, 19 (2006) 127-141. [56] B.P. Leonard, A stable and accurate convective modelling procedure based on quadratic upstream interpolation, Comput. Methods Appl. Mech. Eng., 19 (1979) 59- 98. [57] R.I. Issa, Solution of the implicitly discutised fluid flow equations by operatingsplitting, J. Comput Phys, 62 (1986) 40-65. ! ! ! ! Chapter 4 Biological adhesion – Effect of surface properties Abstract Microchannels fabricated with different biocompatible polymers were utilized to determine the adhesion rate with Escherichia coli (E. coli) suspension. Different polymeric surfaces - namely: cellulose acetate (CA), polydimethylsiloxane (PDMS), polystyrene (PS), polyamide (PA), poly L-Lactide (PLLA) and polyethylene oxide (PEO) - were fabricated and implemented into the microchannel through the developed in-house method along with the xurographic technique. The adhesion rate for all the polymeric surfaces were compared for two different wall shear stresses namely 0.01 Pa and 0.02 Pa. Along with 6 polymeric surfaces, glass was used as a control surface. The surfaces were determined as hydrophilic or hydrophobic based on the surface contact angle obtained by the sessile drop method. Attempts were made to relate adhesion rate with wall shear stress, hydrophobicity and zeta potential. No significant trends were identified, suggesting that for the two shear stresses studied, adhesion is independent of the surface properties. Further studies are necessary, with a larger range of shear stresses. Adhesion on PDMS surface was studied for a range of shear stresses from 0.01 Pa to 2 Pa. Adhesion rate was small for very low shear stress and attained a maximum for 0.2Pa decreasing for larger wall shear stresses. Also adhesion rate obtained for 0.01 Pa and 0.02 Pa were compared with the macroscale study with parallel plate flow chamber. There was not a statistical significant difference between adhesion at different scales for the same wall shear stress. The computational fluid dynamics simulation was used to verify the shear stress in the region of interest in the microchannel. This work was developed to study the adhesion of living cells in polymeric surfaces inserted in microchannels through the technique detailed in chapter 3. CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 74! 4.1 Introduction There are a variety of biomedical applications, for example, arterial implants, catheters, blood pump systems or controlled drug delivery valves, in which chronic infections occur [1]. In these infections, the first step is the adhesion of microorganisms over a conditioning film (saliva, blood, urine etc.) covering the surface. Biofilm formation is the second step, which depends on the Brownian motion, gravitation, diffusion, convection or intrinsic movement of the microorganisms [2]. Biomedical applications use, sometimes, smart materials at small scales for novel functions [3]. Therefore, there is a need to understand the dynamics of biofilm development and development in small scale materials to prevent infections [4]. The adhesion of microorganisms to a surface is also an important study in food process industries and in many chemical and biomedical industrial processes. For over two decades, membrane biofouling in water treatment systems has been an intense research topic [5]. However, a complete understanding of the initial fouling mechanism is still lacking. Several effects on cell adhesion have been studied: different surface materials [6], different bacteria [7], hydrophilicity and hydrophobicity of the surfaces [8], electrical charges of the bacteria [9], bacterial cell wall hydrophobicity [10, 11] and wall shear stresses [12, 13]. Many studies reported the adhesion on different surface materials. A recent one stated that adhesion is governed by a combination of physico-chemical- macromolecular and wall shear force effects [14]. Bacterial adhesion studies on glass slides were performed for shear rates ranging from 30 to 200 s-1 and it was reported that bacterial colonization starts at high wall shear rates and the subsequent monolayer develops at a low shear rate after 1200 hours of experiment [15]. In microchannels, the number of studies about adhesion in polymeric surfaces is scarce. Adhesion of L1210 cells on sulphonated and nonsulphonated micro polymeric surfaces [16] and cell adhesion on poly-methylmethacrylate (PMMA) were reported CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 75! on rough and smooth micro surfaces [17]. Very few adhesion studies have been carried out in surface modified PDMS microchannels [18-20]. Surface modification has paved way to reduce biofouling [21], in particular when applied to microfluidic devices. Sharma et al. [22] experimented two techniques, gasphase modification and wet chemical modification, to develop a stable surface. Pinto et al. [23] modified the PDMS surface using an argon low-pressure plasma technique and reported that the surfaces became non-toxic and slightly haemolytic making them suitable for several biomedical applications. The impact of electrostatic interactions, an often ignored effect, was studied by Loosdrecht et al. [24] to predict initial bacterial adhesion. Apart from the wall shear stress, the zeta potential effect was studied to understand the impact of electrostatic interactions on adhesion in different polymeric surfaces. A similar study was developed by Wang et al. [25] on different glycopolymer surfaces, reporting that lectin-carbohydrate interactions had much more significance on adhesion than electrostatic interactions. A higher adhesion was observed with P. aeruginosa than with Escherichia coli on stiffer glycopolymer surfaces in high ionic strength conditions. Few adhesion studies were done on motile and non motile E. coli, according to the velocity of the suspension flow [26]. Experiments were carried out with microorganisms with and without the presence of flagella and appreciable differences were reported [27]. In this chapter, a study of E. coli adhesion to polymeric surfaces is presented. The experiments were performed at 37 °C and at different wall shear stress conditions. The objective was to find a relationship between wall shear stress and thermodynamic properties, namely hydrophobicity and Zeta potential. The polymeric surfaces were inserted in the microchannels through the technique detailed in Chapter 3 (section 3.2.4). CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 76! 4.2 Materials and methods ! 4.2.1 Selection of polymers A survey on different biocompatible biomedical polymers was carried out having in mind their limitations to be used in microfluidics. The polymers must be transparent after coated to the slides, they should be spincoatable to have uniform thickness and also be able to be retained in the substrate till the end of the experiment and even for more time [28, 29]. Several biomedical polymers usually used in biomedical applications were reviewed and their properties and applications listed. After eliminating the polymers that do not respect the three criteria, nine biocompatible polymers from different fields [30-32] were selected and manufactured. Figure 4.1. Flow chart describing the selection of polymers for the initial adhesion study. The polymers selected were cataloged in terms of applications such as: urinary catheters, urinary bladder, tracheal, etc (Table 4.1). The polymers were prepared and spincoated over a glass slide and durability tests were performed at the shear stresses of the assays [33], only five polymers were suitable to be used: cellulose acetate (CA), polyamide (PA), polydimethylsiloxane (PDMS), poly L-Lactide (PLLA) and polyethylene oxide (PEO). CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 77! Table 4.1 Polymers application Polymer Application Selected references Cellulose acetate (CA) Orthopedic applications due to its strength and stiffness. [34, 35] Polyamide (PA) Woven mesh implant (SupraFOIL, Supramesh, S. Jackson, Inc., Alexandria, VA) [36] Polydimethylsiloxane (PDMS) Micro-transponder system for measuring intra ocular pressure, blood pressure and a system for stimulating nerve cells. PDMS based porous inorganic hybrid materials for good hepatite particle deposition to get a good bond with soft and hard body tissues. [37, 38] Poly L-Lactide (PLLA) Endotine forehead fixation device, endotine midface (Coapt systems, Inc., Palo Alto, CA) [36] Polyethylene oxide (PEO) Micro container for drug delivery applications. Different concentration ranges are used for delivering drug for rectal, ophthalmic, parenteral and percutaneous applications. [38] Polystyrene (PS) (commercially available Petridish) Cell culturing. [39] 4.2.2. Zeta potential and water contact angles for different polymers ! Zeta potential was determined with a Nano Zeta sizer from Malvern Instruments by mixing the dried polymer particles in a solution. Zeta potential is the main index that provides information about the electrostatic repulsion between the polymer surface and the cell assay [40]. When the zeta potential is high, the repulsion force between similarly charged particles is strong and aggregation is avoided. When the zeta potential value is small, the repulsion between similarly charged particles is weak. The zeta potential was calculated using Helmholtz–Smoluchowski equation (1). The water contact angles and the hydrophobicity of the polymers listed in Table 4.1 were obtained as described in Chapter 3 (section 3.3). ζ =4 πµη /D (1) 4.2.4 Wall shear stress from CFD Numerical simulations were made in Ansys Fluent CFD package (version 14.5) to obtain the wall shear stresses in the microchannels. A rectangular microchannel of CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 78! width 450 µm, height 100 µm and length of 5000 µm was the model created by Design Modeller 14.5. The schematic representation of the microchannel is shown in Figure! 4.2. The numerical grid was constructed using Meshing 14.5. The grid was denser in the first half of the channel; 124,154 hexahedral cells uniformly distributed, in counterpart to 94,374 hexahedral cells in the second half. Results were obtained by solving the Navier-Stokes equations for laminar regime using the PISO algorithm, and the QUICK and PRESTO schemes. The no slip boundary condition was considered in all the walls. Figure 4.2. Straight microchannel with dimensions In the simulations, a uniform velocity profile was set at the inlet and the relative pressure was set to zero at the outlet. The inlet velocity ranged between 2.5 ×10-4 and 1.45 ×10-3 m/s, corresponding to a flowrate between 1.13 ×10-11 and 4.34 ×10-12 m3/s and a Reynolds number between 3.6 ×10-2 and 6.22 ×10-2. For the fluid, the properties of water (density and viscosity) at 37 ºC were used. Simulations were made in transient mode, to assure convergence and to capture transient flow structures. For each case, 2s of physical time were simulated with a fixed time step of 10-4 s. Two kinds of microchannels were used in this chapter. The microchannel that was used for studying the initial adhesion and the effect of the flow rate is represented in Figure!4.3 is the microchannel that is used for studying the initial adhesion based on the effect of flow rate. The other microchannel, which was used for studying the initial adhesion and the effect of the polymeric surface, is shown in chapter 3. Simulations with this microchannel were made to understand whether there are large changes in the WSS values before and after the lowered surface. For a particular flowrate, the wall shear stress was obtained by numerical simulation. Figure 4.3 represents the wall shear stress along the rectangular microchannel. The wall shear stress is constant along all the flow direction. A detailed CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 79! representation of the viewing region in Figure 4.3 shows that the wall shear stress is constant in the region where adhesion will be measured. Figure 4.3. Wall shear stress in the microchannel obtained by numerical simulation ! ! 4.2.5. Experimental setup ! The experimental setup for obtaining the adhesion rate in different polymeric surfaces is described in Chapter 3. A schematic representation of the microchannel with the polymeric surface is in Figure!4.4. The cell suspension enters the inlet port, with the help of a syringe pump, at a particular flowrate and, by consequence, a well-defined wall shear stress. The microscope was used to visualize the adhesion in the polymeric surface. CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 86! Table 4.3. Wall shear stresses and correspondent adhesion rates for PDMS microchannels Wall shear stress (Pa) Adhesion rate (nº of cells/(cm2.s)) 0.0104 499.13 0.0208 838.5 0.154 1029.6 0.231 829.7 0.307 510.78 0.346 274.4 0.384 237.88 0.484 288.33 0.75 207.1 1 199.46 2 111.71 Figure 4.8. Fouling rate in a PDMS surface versus wall shear stress applied. Ranges of wall shear stress in human circulatory settings: 1) vena cava, aorta, veins urinary catheter, bladder implant; 2) venules, infra-renal aorta [43] ; 3) arteries; 4) capillaries. It should be stressed that the adhesion for low wall shear stresses grows linearly along the 30 minutes duration of the experiment (Figure 4.9 for 0.0208 Pa) while for high values (Figure 4.10 and Figure 4.11 for 2 Pa) it is linear during a short initial period remaining almost constant after that, so the trendline was fitted for first initial 200 s for the example shown in Figure 4.11. CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 87! Figure 4.9. E. coli bacteria adhesion along experiment for a low wall shear stress (0.0208 Pa). Linear trend during 1200 s. Figure 4.10. E. coli bacteria adhesion along experiment for a high wall shear stress (2 Pa). The adhesion stops after 200 s. Figure 4.11. Trendline fitted up to 200 s for the plot in Figure 4.10 CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 88! 4.3.4 Contact angle effect on adhesion The contact angle effect on cell adhesion is plotted in Figure!4.12 for two different wall shear stresses, 0.01 Pa and 0.02 Pa. The adhesion rates for PA (69.40), PDMS (113.60) and glass (16.40) at 0.02 Pa are higher than at 0.01 Pa while the adhesion in PLLA (88.030) is more or less independent of the wall shear stress applied. A different trend can be observed for CA (65.20), PEO (55.50) and PS (80.80); the adhesion rate is higher for 0.01 Pa than for 0.02 Pa. The diversity of results explains why certain materials are chosen for specific applications. With these different trends there is no clear effect of the contact angle on the cell adhesion. This effect needs to be further studied with more wall shear stresses and different surfaces. Figure 4.12. E. coli bacteria fouling rates on different polymers at two wall shear stresses (0.02 and 0.01 Pa) versus contact angle CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 89! 4.3.5. Effect of hydrophobicity on adhesion The hydrophobicity values of the surfaces calculated in Chapter 3 are reproduced in Table!4.4. The negative values are more hydrophilic in nature. According to the table, PLLA (- 65.32 mJ/m2) is the most hydrophilic material, while glass is the most hydrophilic (27.99 mJ/m2). When the hydrophobicity is high, the cells should have difficulty to adhere to the surface. However, the results in Figure 4.13 do not show this trend. A detail study is needed for further clarification. Table 4.4. Hydrophobicity for different surfaces along with adhesion rates for 0.01 Pa and 0.02 Pa. Surface Hydrophobicity (mJ/m2) 0.01 Pa cells/(cm2s) 0.02Pa cells/(cm2s) GLASS 27.99 595.9 762.2 PEO 0.34 578.9 474.4 CA -36.04 837.6 570.4 PA -37.58 379.5 891.4 PS -49.56 514.4 290.3 PDMS -61.81 446 741 PLLA -65.31 419.2 457.8 Figure 4.13. E. coli bacteria fouling rate versus hydrophobicity of the surface 0! 300! 600! 900! 1200! -80! -50! -20! 10! 40! !Fouling!rate!(No.!of!cells/cm2/s)! Hydrophobicity!(mJ/m2)!! 0.01!Pa! 0.02Pa! CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 90! 4.3.6. Effect of zeta potential on adhesion The zeta potential values obtained for all the polymer surfaces are negative as shows Table!4.5. The E. coli is a Gram-negative bacterium with negative charge. Therefore both have negative charges (E. coli and polymer surface) favoring the repulsion instead the adhesion. On seeing the results in Figure!4.14, where the adhesion level is high, adhesion rate does not seem to have any solid relation with the zeta potential values. Proper further insight into this topic is required to understand adhesion rate in terms of zeta potential. Table 4.5 Zeta potential values for each polymeric surface Polymeric surfaces Zeta potential (mV) CA -23.4 PA -28.0 PDMS -29.3 PLLA -27.9 PEO -11.0 PS -29.6 Glass -37.0 Figure 4.14. E. coli bacteria fouling rate versus zeta potential 0! 300! 600! 900! -40! -30! -20! -10! !Fouling!rate!(No.!of!cells/cm2/s)! Zeta!potential!!(mV)! 0.01!Pa! 0.02Pa! CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 91! 4.3.7 Adhesion rate in micro and macro scales. The adhesion rates in the microchannel were compared with those obtained in a macroscale parallel plate flow chamber (PPFC- the LEPABE group of FEUP has this equipment). The microscopic slides were coated with the polymers and the adhesion rates were analyzed in the PPFC at the same wall shear stress (microchannel) through different flowrates. The volumetric scale-up factor, between PPFC and microchannel, is about 50000x, while the aspect ratio is higher in the microchannel. The average wall shear stress chosen for the comparison was 0.02 Pa. As already referred, approximate shear stress values can be found in the human body at different locations like: urethra [53], uterus [54] and veins [55]. In Figure!4.15, the adhesion rates obtained in the microchannel and in the PPFC are compared. It is observed that, when similar wall shear stresses are used in both scales, similar adhesion rates are obtained whatever the surface. This similarity is an advantage, since it allows the choice of the scale experiment depending on the expertise and equipment available. Micro flow systems can be used taking advantage of lower hold-up volumes or macro flow systems can be selected in order to obtain a higher biofilm mass, which can be used for further biochemical analysis [33]. The highest adhesion rate is in PA for both scales and the lowest in PLLA also for both scales. The highest adhesion rate was expected to be found in a hydrophobic surface and the lowest in a hydrophilic surface (glass [56]). However the results do not follow this trend as explained before. CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 92! Figure 4.15. E. coli bacteria adhesion rates on PA, glass, PDMS, CA and PLLA obtained in the microchannel (black bars) and in the PPFC (white bars). Error bars shown for each surface represent the standard deviation from three independent experiments [33] 4.4. Conclusions and future work. The adhesion study performed in a PDMS surface for wall shear stresses ranging from 0.01 Pa to 2 Pa shows a consistent trend. After attaining a peak near 0.2 Pa, the adhesion rate gradually falls down up to 0.4 Pa. For higher shear stress values, the adhesion rate decreases, converging to almost zero. This data can be used to understand as how the PDMS surface would get fouled while operating in an implant or in any micro medical device. A consistent correlation between bacterial adhesion rates and material surface properties (hydrophobicity and zeta potential) was not found. Further studies should be performed to better understand the adhesion on a thermodynamic scale. The polarity of the surfaces should also be considered to get an insight into the adhesion phenomenon. Also a database should be created for each biomaterial. The biofouling rate comparison between micro and macro flow systems showed interesting results. The observed similarity between the adhesion rates is an added CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 93! advantage, as it allows the choice of the scale experiment depending on the expertise and equipment available on the lab. References. [1] T. Bjarnsholt, The role of bacterial biofilms in chronic infections, APMIS, 121 (2013) 1-58. [2] R. Bos, H.C. van der Mei, H.J. Busscher, Physico-chemistry of initial microbial adhesive interactions - its mechanisms and methods for study, Fems Microbiol Rev, 23 (1999) 179-230. [3] F. Greco, V. Mattoli, Introduction to Active Smart Materials for Biomedical Applications, in: G. Ciofani, A. Menciassi (Eds.) Piezoelectric Nanomaterials for Biomedical Applications, Springer Berlin Heidelberg, 2012, pp. 1-27. [4] G.A. Truskey., F. Yuan., D.F. Katz, Transport Phenomena in Biological Systems, 2004. [5] H.S. Vrouwenvelder, J.A.M. van Paassen, H.C. Folmer, J.A.M.H. Hofman, M.M. Nederlof, D. van der Kooij, Biofouling of membranes for drinking water production, Desalination, 118 (1998) 157-166. [6] U. Hersel, C. Dahmen, H. Kessler, RGD modified polymers: biomaterials for stimulated cell adhesion and beyond, Biomaterials, 24 (2003) 4385-4415. [7] J.H. Lee, G. Khang, J.W. Lee, H.B. Lee, Interaction of Different Types of Cells on Polymer Surfaces with Wettability Gradient, Journal of Colloid and Interface Science, 205 (1998) 323-330. [8] T. Horbett, J. Waldburger, B. Ratner, A. Hoffman, Cell adhesion to a series of hydrophili–hydrophobic copolymers studies with a spinning disc apparatus, Journal of biomedical materials research, 22 (1988) 383-404. CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 94! [9] M.A. Lopes, F.J. Monteiro, J.D. Santos, A.P. Serro, B. Saramago, Hydrophobicity, surface tension, and zeta potential measurements of glass-reinforced hydroxyapatite composites, Journal of Biomedical Materials Research, 45 (1999) 370- 375. [10] M.C. van Loosdrecht, J. Lyklema, W. Norde, G. Schraa, A.J. Zehnder, The role of bacterial cell wall hydrophobicity in adhesion, Appl Environ Microbiol, 53 (1987) 1893-1897. [11] R. Oliveira, J. Azeredo, P. Teixeira, A. Fonseca, The role of hydrophobicity in bacterial adhesion, (2001). [12] J.E. Duddridge, C. Kent, J. Laws, Effect of surface shear stress on the attachment of Pseudomonas fluorescens to stainless steel under defined flow conditions, Biotechnology and bioengineering, 24 (1982) 153-164. [13] P.F. Davies, A. Robotewskyj, M.L. Griem, Quantitative studies of endothelial cell adhesion. Directional remodeling of focal adhesion sites in response to flow forces., J Clin Invest, 93 (1994) 2031-2038. [14] M.G. Katsikogianni, Y.F. Missirlis, Bacterial adhesion onto materials with specific surface chemistries under flow conditions, Journal of materials science. Materials in medicine, 21 (2010) 963-968. [15] T. Paris, S. Skali-Lami, J.C. Block, Effect of wall shear rate on biofilm deposition and grazing in drinking water flow chambers, Biotechnology and bioengineering, 97 (2007) 1550-1561. [16] J. Dobkowski, R. Kolos, J. Kamiński, H.M. Kowalczyńska, Cell adhesion to polymeric surfaces: experimental study and simple theoretical approach. , J Biomed Mater Res, 47 (1999) 234-242. CHAPTER(4((-(BIOLOGICAL(ADHESION(–(EFFECT(OF(SURFACE(PROPERTIES! ! J.!Ponmozhi!–!PhD!Thesis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! 95! [17] M. Lampin, R. Warocquier-Clérout, C. Legris, M. Degrange, M.F. Sigot- Luizard, Correlation between substratum roughness and wettability, cell adhesion, and cell migration, Journal of Biomedical Materials Research, 36 (1997) 99-108. [18] H. Makamba, J.H. Kim, K. Lim, N. Park, J.H. Hahn, Surface modification of poly (dimethylsiloxane) microchannels, Electrophoresis, 24 (2003) 3607-3619. [19] H. Lu, L.Y. Koo, W.M. Wang, D.A. Lauffenburger, L.G. Griffith, K.F. Jensen, Microfluidic Shear Devices for Quantitative Analysis of Cell Adhesion, Anal Chem, 76 (2004) 5257-5264. [20] Z.L. Zhang, C. Crozatier, M. Le Berre, Y. Chen, In situ bio-functionalization and cell adhesion in microfluidic devices, Microelectronic Engineering, 78-79 (2005) 556- 562. [21] J. Zhou, D.A. Khodakov, A.V. Ellis, N.H. Voelcker, Surface modification for PDMS-based microfluidic devices, Electrophoresis, 33 (2012) 89-104. [22] M.D. V. Sharma, S.M. Shivaprasad, S.C. Jain, Surface characterization of plasma-treated and PEG-grafted PDMS for micro fluidic applications, Vacuum, 81 (2007) 1094-1100. [23] S. Pinto, P. Alves, C.M. Matos, A.C. Santos, L.R. Rodrigues, J.A. Teixeira, M.H. Gil, Poly(dimethyl siloxane) surface modification by low pressure plasma to improve its characteristics towards biomedical applications, Colloids and Surfaces B: Biointerfaces, 81 (2010) 20-26. [24] v.L.M. C., L. J., N. W., S. G., Z.A. J, Electrophoretic mobility and hydrophobicity as a measured to predict the initial steps of bacterial adhesion., Appl. Environ. Microbiol., 53 (1987) 1898-1901. [25] Y. Wang, R. Narain, Y. Liu, Study of Bacterial Adhesion on Different Glycopolymer Surfaces by Quartz Crystal Microbalance with Dissipation, Langmuir, 30 (2014) 7377-7387. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 102 microchannel is more or less favorably to fouling. So, in this chapter, fouling in microfluidic devices with different microchannel constrictions is monitored to understand the initial adhesion mechanisms in different locations. Moreover, it is also important to understand how cells behave and how the formation of biofilms can hinder the normal functions of the device. This knowledge is important to know how long a system can work without maintenance or when substitution is indispensable. Microfluidic devices are useful to study cell adhesion in a controllable dynamic environment simulating in vivo conditions. The main conclusions in microchannels can be extrapolated, adopting similarity rules, to larger scale devices. A microfluidic system can be designed to provide information about fouling rates on specific materials, antibiotics minimum dosage [2], quorum sensing [3] or on the relation between shear stress and biofouling [4, 5]. Some devices have been designed with the specific intention of studying biofilm related problems. Meyer et al. [6] developed a platform for monitoring the biofilm formation by measuring its optical density, which has the potential to be used as a standalone biochip for testing cleaning methods. Ghodssi et al. [7] developed microfluidic systems for optical and surface acoustic waves based detection and quantification of biofilms. Kim et al. [8] developed a device to determine the minimum concentration of antibiotics for biofilm eradication using generated antibiotics gradients. Zhang et al. [9] developed a procedure to bio functionalize a micro channel wall with antibodies to promote cell adhesion. Valiei et al. [10] used a microfluidic device to study the development of streamers and their role on biofilm growth. They found that the formation and morphology of the streamers depended on the flow rate. Rusconi et al. [11] observed the formation of streamers in microfluidic devices with curved-section channels. They proposed a relation between streamers development and secondary flows in the corners. These are examples on how microfluidics can be used for detailed observation of biofilm evolution and also situations in which fluid dynamics is relevant for biofilms formation. Biomimetic design approach has been proposed for developing optimized geometries [12-15]. Networks following Murray's law are known to have a constant shear rate in all the branches. Barber et al. [13] proposed the application of the CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 103 Murray's law statement of constant wall shear stress, to constant depth rectangular cross section channels made by standard planar microfabrication techniques like photolithography. This design, when utilized in systems with fouling, reduces the risk of the appearance of critical points and the subsequent clogging of the channel. Emerson and Barber [15] extended this analysis to non-Newtonian fluids, concluding that the optimal design for a power law fluid requires detailed numerical simulations, since a simple analytical approach is not possible. As described in chapter 4, the cell adhesion over a surface was studied by several authors along the years from 1970’s: experimentally [16, 17], numerically, considering just a single spherical cell adhered to a surface [18], and also theoretically [19]. Very few papers were published about cell adhesion in microchannels and only some of them describe the methods to reduce fouling in microchannels. Hang Lu et al. [20] studied cell attachment and detachment in a 25 x 500 µm microchannel. They applied relatively high shear stresses (160 Pa) using a multi-shear device and constant flow rate. Fuhr et al. [21] studied the reduction of cells and microparticles adhesion through an electric field created by ultra-microelectrodes fabricated by electron-beam lithography on silicon wafers, to create bio-repellent surfaces. They proposed two experimental techniques: one passing the erythrocyte cell suspension through the microchannel that had the electrode array for several hours; the second by growing fibroblasts under constant electric field. They showed that there was no cell adhesion on the substrate with amplitudes between 0.5 and 2 volt for both experiments performed. In another cell adhesion study in a microchannel [22], a method was described to measure the adhesion forces of a bacteria possessing different types of pili’s. Polymer stencils that function as scaffolds to control cell adhesion inside microchannels were used to measure biofilm formation both in static and flow conditions [23]. Vijay et al. [3] tried to predict a physical model of the biofilm grown in a microchannel with Pseudomonas aeruginosa. A similar initial adhesion study was done by Velraeds et al. [24] with different uropathogenic bacteria, fouling during 4 h a silicone rubber surface, to understand urinary tract infection in hospitals. Liu et al. [25] investigated the effects of environment, micro geometry and flow on bacterial adhesion. They showed that, the adhesion is influenced by cell culture time, microchannel geometry and cell suspension flowrate. According to the study, the adhesion increases in complex geometries. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 104 In this chapter, a study about the effect of microchannel constrictions on cell adhesion is presented. It is known that some spots in channel networks can be critical points prone to fouling and clogging. In this study, initial adhesion in constrictions, multiple constrictions and channels with variable cross sections are analysed and adhesion rates are related to local wall shear stresses and flow patterns. 5.2. Materials and Methods 5.2.1. Microchannel designs ! Three different constrictions were used with different constriction lengths as shown in Figure 5.1 and Table 5.1. The constriction lengths (!!) are 5000 µm, 2000 µm and 500 µm with entrance length (!!) of 5000 µm for all the channels. The design was used to identify areas of higher adhesion: before, after or in the constriction at a pre-tested, high adhesion over a short period of time at particular shear stress. The second geometry, Figure 5.2, is a converging channel followed by a sudden expansion. The third design, is a channel with multiple constrictions within a short span of 5000 µm distance. Figure 5.1. Sudden constriction channel design (not to scale) Table 5.1. Dimensions of the channels with a sudden contraction ! Constriction!! AA! BB! !!! (µm)! !!! (µm)! H" (µm)! W" (µm)! H" (µm)! W" (µm)! A! 94.4! 88.5! 94.4! 8.54! 5000! 5000! B! 94.4! 86.4! 94.4! 10.5! 5000! 2000! C! 94.4! 92.6! 94.4! 13.6! 5000! 500! CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 105 Figure 5.2. Converging channel design (not to scale) Figure 5.3. Multiple sudden constriction channel design (not to scale) The cell suspension preparation was discussed in detail in chapter 3 and chapter 4, while the experimental set up (Figure 5.5) was described in detail in chapter 3 and will be briefly explained in section 5.2.4. 5.2.2 Flow conditions In this study, the flow is laminar, viscous forces dominate, and so Reynolds number is small. Depending on the channel dimensions, measured after fabrication, the flow rate will be set through the constant pre-established value of the shear stress at the wall. The shear stress at the wall, w τ , is directly proportional to the flow rate Q in the microchannel, to the fluid viscosity  (the cell suspension has physical properties similar to water, a Newtonian incompressible fluid) and inversely proportional to the width and to the square height of the channel. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 106 2 6 wh Q w µτ = (1) Equation 1 is an approximation, valid when w is significantly larger than h. The values obtained by this equation were used as the base to design the experiments and will be referred as “nominal wall shear stress”. More accurate values, obtained by computational fluid dynamics, will also be presented.The flow regime was determined by calculating the Reynolds number based on the width and height of the microchannel: (2) The shear stress values selected for the different microchannels are presented in Table 5.2, along with the flow rates (calculated from equation 1), channel width and Reynolds number (always less than 1). The shear stresses values will be compared with data from the numerical simulations. Cell adhesion was obtained for four different nominal shear stresses. Nominal shear stresses are evaluated in region of interest (in the middle of the microchannel length) in all the five different channels with different types of constrictions, as tabulated in Table 5.2. These varied types of microchannels were designed for the purpose of studying the minimal fouling areas in different zones (especially near constrictions) of the microchannel with shear stresses ranging from 0.1 to 1 Pa. These shear stresses were selected for the study, because the adhesion rate, in these conditions, is high as shown in Table 5.3 (conclusions from chapter 4, section 4.3.3). Re = ρ Q (W+H) µ CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 107 Table 5.2 Shear stress at the region of interest and flow rates values employed in studies with different microchannels Nominal!wall! shear!stress! (Pa)! Geometry! Flow!rate! (m3/s)! Upstream! channel! width!(µm)! Reynolds! number! 0.1! constriction! 1.71!×!10-11! 79.7! 0.31! 0.2! constriction! 3.42!×!10-11! 79.7! 0.62! converging! microchannel! 8.58!×!10-11! 200! 0.62! multiple!constrictions! 3.47!×!10-11! 80.8! 0.62! 0.3! constriction! 5.13!×!10-11! 79.7! 0.92! 1! constriction! 1.71!×!10-10! 79.7! 3.08! converging! microchannel! 4.29!×!10-10! 200! 3.08! multiple!constrictions! 1.73!×!10-10! 80.8! 3.08! A list of adhesion rates corresponding at different wall shear stresses, from chapter 4, are listed in Table 5.3 for an easy understanding of shear stress relation to adhesion rate. In this range of wall shear stress, the adhesion rate decreases as the wall shear stress increase. Table 5.3 Adhesion rates and respective shear stresses (from chapter 4) Wall!shear! stress!(Pa)! Adhesion!rate!(nº!cells/! cm2.s)! 0.154! 1029.6! 0.231! 829.7! 0.307! 510.78! 1! 199.46! 5.2.3 Numerical simulations The flow in the microchannels was simulated to determine the wall shear stresses. A section of the microchannel, around the visualization region, was selected for simulation domain (Figure 5.4). CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 108 Figure 5.4. 2D representation of the numerical domains: (a) constriction geometry; (b) multiple constrictions; (c) converging microchannel. ! Numerical simulations were made in Ansys Fluent CFD package (version 14.5). Models of the microchannels were built in Design Modeler 14.5 and were discretized into a grid by Meshing 14.5. Results in the laminar regime were obtained by solving the Navier– Stokes equations. The velocity–pressure coupled equations were solved by PISO algorithm (Issa 1986), QUICK scheme (Leonard 1979) [26] was used for the discretization of the momentum equations. The simulation is carried out with the same method as described in the previous chapter. 5.2.4 Experimental set-up The experimental set-up was the same described as in Chapters 3 and 4. The E. coli suspension was used to fill the syringe and the syringe pump controlled the flow rate. The working temperature was monitored with the RTD sensor throughout the experiment. Figure 3.8 in Chapter 3 and Figure 4.2 in Chapter 4 describe pictorially the set-up. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 109 Figure 5.5. Schematic representation of monitoring the whole length of the microchannel with the microscope objective for every 500 µm interval. The adhesion study was performed for 30 mins following the same procedure described in Chapter 4. Additionally for this chapter at the end of the experiment, i.e. after 30 mins of adhesion study, the whole length of the microchannel was observed with the microscope. Images were captured for every 500 µm interval length of the microchannel by moving the microscope objective as shown in the Figure 5.5 throughout the whole length of the microchannel. 5.3 Results and Discussion 5.3.1. Constrictions 5.3.1.1. Wall shear stress effect The adhesion along the channel constrictions was monitored for all the microchannel designs shown in Figure 5.1, 5.2 and 5.3 Unlike the straight rectangular channels where adhesion was uniform along all its length (Chapter 3 - section 3.9.4), adhesion is not uniform along the channels with constrictions. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 110 Figure 5.6. Adhesion along the 2000 µm constriction microchannel at different shear stresses The effect of shear stress was studied for a microchannel with a 2000 µm constriction (Table 5.1). Four wall shear stress values were studied; 0.1, 0.2, 0.3 Pa and 1 Pa. In Figure 5.6, the xaxis refers to the length of the microchannel (in µm) and the y-axis to the number of cells adhered per square centimeter. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 111 Figure 5.7. Microscopic image of 2000 µm constriction at different spots along its length for 0.2 Pa shear stress: (a) before the constriction (uniform adhesion); (b) start of the constriction (sudden contraction - sudden increased adhesion); (c) end of the constriction (decreased adhesion); (d) after the constriction (sudden expansion – highly decreased adhesion); (e) location of a,b,c and d adhesion zones along the 2000 µm constriction microchannel. The adhesion gradually decreased as the wall shear stress increases from 0.1 Pa to 1 Pa. This was clearly noticeable in the upstream, constriction and downstream regions of the microchannel. For the shear stress of 1 Pa, the adhesion is very low, as expected by observation of the adhesion rates in CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 118 In Figure 5.15a the adhesion along the length is represented in simultaneous with the width of the microchannel. The adhesion decreases as long as the width decreases and, in the sudden expansion region, it takes some length to attain the uniform adhesion value. In Figure 5.15 b, the adhesion is represented in simultaneous with the wall shear stress at each location along the length. In very high adhesion regions, the wall shear stress is low and vice-versa. This behaviour was justified in the previous chapter (Chapter 4). Figure 5.16. Converging microchannel for 1 Pa: (a). Adhesion along the length of the converging microchannel for 1 Pa at the inlet, (b) Adhesion with respective channel width at each location along the length, (c) Adhesion with wall shear stress at each location along the length. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 119 Figure 5.16 a and b shows the experimental data for wall shear stress 1 Pa at the inlet. As the width of the channel decreases, the adhesion also decreases and its level is very low at the end of the converging channel. Contrasting to 0.2 Pa experiments, the adhesion in the sudden expansion region recovers in a short length and maintains the same low rate in the downstream region. These differences can be clearly seen in the images of Figure 5.17 and Figure 5.18.The upstream adhesion decreases slowly until the sudden expansion zone for 0.2 Pa experiments (Figure 5.17 a) and the downstream adhesion, after the convergence, has a gradual increase until the end of the microchannel (Figure 5.17 b). Figure 5.17. Microscopic image for the converging microchannel at 0.2 Pa: (a) End of the converging region; (b) Beginning of the expansion region after the convergent; (c) Location of a and b on zones along the converging microchannel For the 1 Pa experiments, the upstream adhesion dropped to a greater extent but at the end of the converging channel (Figure 5.18 a) a significant number of cells appear to block the passage. This phenomenon does not occur in the multiple sudden constrictions or in the three different constriction microchannels analysed, where high CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 120 shear stresses induce low or null adhesion. Downstream the convergence region the adherence does not increase (Figure 5.18 b). Figure 5.18. Microscopic image for the converging microchannel at 1 Pa: (a) End of the converging region; (b) Beginning of the expansion region after the convergent). (c) Location of a and b zones along the converging microchannel 5.4 CFD An example of the flow in a constriction is represented in Figure 5.19 and Figure 5.20. The flow is in the laminar regime, with Reynolds number lower than 1 in all cases. The flow is characterized low lengths of development and symmetry between the flow entering and leaving the constriction. High velocity and high shear stress is observed in the constriction. Since the lengths of flow development are much shorter that 500 µm no significant difference is observed as the constriction length increases. Similar conclusions can be taken, even for the larger flow rates studied. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 121 Figure 5.19. Magnitude of the velocity in a 500 µm long constriction. The flow rate is 3.42 × 10-11 m3s-1. The constriction width is 10 µm and the channel depth is 100 µm. Figure 5.20. Wall shear stress in a 500 µm long constriction. The flow rate is 3.42 × 10-11 m3s-1. The constriction width is 10 µm and the channel depth is 100 µm. The nominal wall shear stress in the upstream channel is 0.2 Pa. The wall shear stress calculated by the numerical method in the upstream channel is 0.31 Pa. The wall shear stress in the constriction is 19.2 Pa. Table 5.4 shows the wall shear stress for different flow rates in two locations in the channel, the upstream channel and the constriction. A significant correction must be introduced in the nominal wall shear stress. The real wall shear stress is approximately 50% above the nominal wall shear stress. The wall shear stress in the constriction is approximately 40 times the wall shear stress in the upstream channel. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 122 Table 5.4. Wall shear stresses in a microchannel with a 500 µm long constriction. The constriction width is 10 µm and the channel depth is 100 µm. Q!(m3s-1)! τ*(Pa)! τc*(Pa)! Nominal! Numerical! 1.71!×!10-11! 0.1! 0.155! 6.8! 3.42!×!10-11! 0.2! 0.31! 13.65! 5.13!×!10-11! 0.3! 0.465! 20.5! 1.71!×!10-10! 1! 1.55! 68.1! Figures 5.21 and 5.22 show the results obtained for the converging microchannel. As can be observed velocity and shear rate increase significantly along the channel, as was already predicted by the theoretical approach. Figure 5.21. Magnitude of the velocity in a converging channel. The flow rate is 8.58 × 10-11. m3s-1 Figure 5.22. Wall shear stress in the converging channel. The flow rate is 8.58 × 10-11. m3s-1 Figures 5.23 and Figure 5.24 show the results obtained for the multiple constrictions microchannel. As can be observed, the flow is fully developed in the constrictions but the wide channels have a length insufficient to reach developed flow. For this reason the wall shear stress along the centreline of the wide sections is higher than in a longer channel with the same width. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 123 Figure 5.23. Velocity magnitude in a section of a multiple constrictions microchannel. The flow rate is 3.42 × 10- 11 m3s-1. The constriction width is 10 µm and the channel depth is 100 µm. max is 13.8 Figure 5.24. Velocity magnitude in a section of a multiple constrictions microchannel. The flow rate is 3.42 × 10- 11 m3s-1. The constriction width is 10 µm and the channel depth is 100 µm. 5.5 Influence of wall shear stress on adhesion rates The plot for local wall shear stress in different microchannels is plotted with their corresponding adhesion rates in Figure 5.25. This plot is the extension of Figure 4.7. in chapter 4 where adhesion rate in a PDMS surface is plotted for different wall shear stress The Adhesion rate is calculated at different locations along the length of the microchannels. The plot 1, 2, 3, 4 mentioned in Figure 5.25 is adapted from Figure CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 124 5.6, Figure 5.10, Figure 5.12 respectively for the presenting the points A to I to calculate the adhesion rates at those locations for a period of 30 minutes. Figure 5.25. Fouling rates at different locations as mentioned in 1,2,3,4 plots (location of A to I points) with their local wall shear stress. 1- 2000 µm constriction microchannel; 2- 5000 µm constriction for a nominal shear stress of 0.2 Pa; 3 and 4 – multiple constrictions. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 125 The values of adhesion rates with their local wall shear stress for the corresponding nominal wall shear stress is tabulated for all the location points in Figure 5.25 are tabulated in Table 5. Table 5.5. Adhesion rates with local and nominal wall shear stress for the points mentioned in Figure 5.25. Point! Nominal!WSS! Local!WSS! Adhesion!rate!(No.! of!cells/cm2/s)! A! 0.1! 6.8! 488.61! B! 0.2! 13.65! 284.65! C! 0.3! 20.5! 39.63! D! 1! 68.1! 5.99! E! 0.2! 13.65! 95.49! F! 0.2! 13.65! 0! G! 0.2! 13.65! 1815.99! H! 0.2! 13.65! 111.6! I! 1! 68.1! 25.32! The results in Figure 5.25 show that local adhesion rates in constriction deviate significantly from the curve obtained for a straight channel suggesting that other factors other than local wall shear stress is influencing cell adhesion. The following factors may have a role in adhesion and further research is necessary: • Mass transport effects when the fluid suddenly accelerates; • Wall effects when the size of the channel becomes similar to the size of the cells; • 3D flow effect 5.6. Conclusions ! In this chapter, an adhesion study in microchannels with different geometries was performed. The relationship between wall shear stress and cross sectional area of the channel was explored. The results show that adhesion in all of the three sets of microchannel geometry designs analysed follow a similar adhesion pattern. However, an independent investigation should always be carried out to explore singularities of each geometry design. Manipulating the geometry, and by consequence the wall shear stress, it is possible to control cell adhesion, an event with great potential for many biomedical applications. Using a microchannel with multiple constrictions, it is also possible to work on both, high and low wall shear stress ranges, and to monitor the adhesion phenomenon in these two situations. Depending on the application, this CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 126 work can be extended to different bacterial cell based assays at a pre-required temperature, which also plays an important role in cell adhesion. The cell deformation and alignment along different constrictions during the flow will be an interesting future prospective study, as it would help to understand the development and behaviour of many unwanted diseased cells in the human body. References [1] B. Thonon, S. Grandgeorge, C. Jallut, Effect of geometry and flow conditions on particulate fouling in plate heat exchangers, Heat Transfer Eng, 20 (1999) 12-24. [2] P. Sun, Y. Liu, J. Sha, Z. Zhang, Q. Tu, P. Chen, J. Wang, High-throughput microfluidic system for long-term bacterial colony monitoring and antibiotic testing in zero-flow environments, Biosensors and Bioelectronics, 26 (2011) 1993-1999. [3] V. Janakiraman, D. Englert, A. Jayaraman, H. Baskaran, Modeling growth and quorum sensing in biofilms grown in microfluidic chambers, Annals of biomedical engineering, 37 (2009) 1206-1216. [4] J.S. Salek MM1, Martinuzzi RJ., The influence of flow cell geometry related shear stresses on the distribution, structure and susceptibility of Pseudomonas aeruginosa 01 biofilms., Biofouling, 25 (2009) 711-725. [5] C.C. Chan, P.R. Berube, E.R. Hall, Relationship between types of surface shear stress profiles and membrane fouling, Water Res, 45 (2011) 6403-6416. [6] M.T. Meyer, V. Roy, W.E. Bentley, R. Ghodssi, Development and validation of a microfluidic reactor for biofilm monitoring via optical methods, J Micromech Microeng, 21 (2011) 054023. [7] R. Ghodssi, M.T. Meyer, Y.W. Kim, Microsystems for sensing and characterization of bacterial biofilms, in: Sensors, 2013 IEEE, IEEE, 2013, pp. 1-6. CHAPTER(5(–(BIOLOGICAL(FOULING(–(CHANNEL(GEOMETRY(EFFECT(((((( J.Ponmozhi*–*PhD*Thesis* 127 [8] K.P. Kim, Y.-G. Kim, C.-H. Choi, H.-E. Kim, S.-H. Lee, W.-S. Chang, C.-S. Lee, In situ monitoring of antibiotic susceptibility of bacterial biofilms in a microfluidic device, Lab Chip, 10 (2010) 3296-3299. [9] Z.L. Zhang, C. Crozatier, M. Le Berre, Y. Chen, In situ bio-functionalization and cell adhesion in microfluidic devices, Microelectronic Engineering, 78-79 (2005) 556- 562. [10] A. Valiei, A. Kumar, P.P. Mukherjee, Y. Liu, T. Thundat, A web of streamers: biofilm formation in a porous microfluidic device, Lab Chip, 12 (2012) 5133-5137. [11] R. Rusconi, S. Lecuyer, L. Guglielmini, H.A. Stone, Laminar flow around corners triggers the formation of biofilm streamers, Journal of The Royal Society Interface, 7 (2010) 1293-1299. [12] M.S. Razavi, E. Shirani, M. Salimpour, Development of a general method for obtaining the geometry of microfluidic networks, AIP Advances, 4 (2014) 017109. [13] R.W. Barber, D.R. Emerson, Optimal design of microfluidic networks using biologically inspired principles, Microfluid Nanofluid, 4 (2008) 179-191. [14] R.W. Barber, D.R. Emerson, Biomimetic design of artificial micro-vasculatures for tissue engineering, Altern Lab Anim, 38 (2010) 67-79. [15] D.R. Emerson, R.W. Barber, Designing Efficient Microvascular Networks Using Conventional Microfabrication Techniques, in: ASME 2009 Second International Conference on Micro/Nanoscale Heat and Mass Transfer, American Society of Mechanical Engineers, 2009, pp. 159-167. [16] M.B. Lawrence, C.W. Smith, S.G. Eskin, L.V. McIntire, Effect of venous shear stress on CD18-mediated neutrophil adhesion to cultured endothelium, Blood, 75 (1990) 227-237.