Control and readout electronics for miniaturized temperature sensors integrated with an organ-on-a-chip

Uni e sidade do Minho Escola de Engenha ia Ca los Manuel Lima Oli ei a Con ol and eadou elec onics o minia u ized empe a u e senso s in eg a ed wi h an o gan-on-a-chip Ma ço de 2022 UMinho | 2022 Ca los Oli ei a Con ol and eadou elec onics o minia u ized empe a u e senso s in eg a ed wi h an o gan-on-a-chip Ca los Manuel Lima Oli ei a Con ol and eadou elec onics o minia u ized empe a u e senso s in eg a ed wi h an o gan-on-a-chip Ma ço de 2022 Ca los Manuel Lima Oli ei a Con ol and eadou elec onics o minia u ized empe a u e senso s in eg a ed wi h an o gan-on-a-chip Disse ação subme ida na Uni e sidade do Minho pa a ob enção do g au de Mes e em Engenha ia Ele ónica Indus ial e de compu ado es T abalho e e uado sob a o ien ação do(s) P o esso a D .ª G aça Minas e D .° Paulo Sousa Ma ço de 2022 iii DIREITOS DE AUTOR E CONDIÇÕES DE UTILIZAÇÃO DO TRABALHO POR TERCEIROS Es e é um abalho académico que pode se u ilizado po e cei os desde que espei adas as eg as e boas p á icas in e nacionalmen e acei es, no que conce ne aos di ei os de au o e di ei os conexos. Assim, o p esen e abalho pode se u ilizado nos e mos p e is os na licença abaixo indicada. Caso o u ilizado necessi e de pe missão pa a pode aze um uso do abalho em condições não p e is as no licenciamen o indicado, de e á con ac a o au o , a a és do Reposi ó iUM da Uni e sidade do Minho. Licença concedida aos u ilizado es des e abalho A ibuição-NãoCome cial-SemDe i ações CC BY-NC-ND h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/ i ACKNOWLEDGEMENTS I would like o hank P o esso G aça Minas o he good o ien a ion and oppo uni y o be a pa o his p ojec . To P o esso Paulo Sousa, a big hanks o all he guidance, knowledge, disponibili y, and dedica ion o his o ien a ion. Also, hank Tiago Ma os o all he suppo in he elec onics and assembling o he ci cui and all he pa ien and disponibili y ha you had h oughou his p ojec . To my pa en s, hanks o allowing me o ha e an educa ion and o all he pa ience on his ough pa h, all I achie ed un il now, and wha I am se o achie e in he u u e is all hanks o you e o s and ad ice. To my sis e , hank you o all he ad ice and o being always he e o me when I needed i . You a e an example o me, someone o look up o. To Ma ga ida, hank you o all he help, ad ice and o always being he e o me. To all my iends, all he gang, you know who you a e, hanks o all he help you ga e me in hese se en yea s, I could no ha e made i wi hou you all. I is a pleasu e o call you amily, we wen h ough e e y hing oge he and we made i , le s celeb a e. To all he Re ood gang, he olun ee wo k con ibu ed di ec ly o wha I am oday and b ough he mos awesome people o my li e, hank you all, le s keep going. Thank you o all o he unnamed pe sons who somehow con ibu ed o wha I am oday. Finally, a big hank you o you B uno. I am hank ul ou pa h came oge he . Wha e e you a e, you a e my s eng h and he ligh ha guides me h ough my pa h, his one is o you dude. This wo k esul s pa ially o he p ojec NORTE-01-0145-FEDER-029394, RTChip4The anos ics, suppo ed by P og ama Ope acional Regional do No e - No e Po ugal Regional Ope a ional P og amme (NORTE 2020), unde he PORTUGAL 2020 Pa ne ship Ag eemen , h ough he Eu opean Regional De elopmen Fund (ERDF) and by Fundação pa a a Ciência e Tecnologia (FCT), IP, p ojec e e ence PTDC/EMD-EMD/29394/2017. i DECLARAÇÃO DE INTEGRIDADE Decla o e a uado com in eg idade na elabo ação do p esen e abalho académico e con i mo que não eco i à p á ica de plágio nem a qualque o ma de u ilização inde ida ou alsi icação de in o mações ou esul ados em nenhuma das e apas conducen e à sua elabo ação. Mais decla o que conheço e que espei ei o Código de Condu a É ica da Uni e sidade do Minho. ii RESUMO De aco do com um es udo ealizado pelo ins i u o de bioma e iais e engenha ia biomédica da Uni e sidade de To on o nos Es ados Unidos da Amé ica, 4 em cada 1000 pacien es so e am de e ei os ad e sos p o ocados po á macos. Es e p oblema su ge de ido ao ac o de o es e de á macos se em ealizados em animais e, pos e io men e em humanos. A u ilização dos o gan-on-a-chip aumen am a e icácia quando se inicia os es es em humanos po que são mais ep esen ia i os do que os es es em animais. A ecnologia o gan-on-a-chip (OOC) su giu com o obje i o de se possí el eplica aspe os impo an es da isiologia humana e des e modo pode supe a as limi ações dos p ocedimen os adicionais, aumen ando assim a segu ança e e icácia de quem os oma. Apesa da ecnologia OOC o nece uma sé ie de an agens compa a i amen e com as écnicas con encionais, es a ainda ca ece de sis emas pa a moni o ização de mul ipa âme os capazes de o nece in o mações à mic oescala du an e os es es de cul u a de células bem como na es agem de no os á macos. Nes e sen ido, o na-se mui o impo an e o desen ol imen o de mic ossenso es in eg ados em sis emas mic o luídicos pa a moni o iza os di e sos pa âme os celula es de modo a pe cebe que e ei os os á macos p o ocam nas células. Pa âme os como empe a u a, oxigénio, pH, ní el de nu ien es, en e ou os, são a iá eis de in e esse pa a a pe ceção global do me abolismo c iado pela combinação de á macos com células de múl iplos ó gãos. Assim, o abalho desen ol ido nes a disse ação em como obje i o es uda e desen ol e a ele ónica de lei u a e a uação de um mic ossenso de empe a u a, baseado em esis ance empe a u e de ec o (RTD), que se á in eg ado na ecnologia OOC. O sis ema desen ol ido pa a além de o nece po abilidade e baixo cus o, pe mi e ealiza lei u a de á ios mic ossenso es com uma sensibilidade de 15mV/0.1°C. Foi ambém implemen ada uma componen e g á ica que pe mi e ao u ilizado seleciona o senso e acompanha os esul ados em empo eal. Pala as-Cha e: Mic ossenso es de empe a u a, O gan-on-a-chip, esis ance empe a u e de e o , sis emas mic o luídicos. iii ABSTRACT Acco ding o a s udy conduc ed by he Ins i u e o Bioma e ials and Biomedical Enginee ing a he Uni e si y o To on o in he Uni ed S a es o Ame ica, 4 ou o e e y 1,000 pa ien s su e ed om d ug- ela ed ad e se e ec s. This p oblem a ises because he d ug es mos ly pe o med on animals and, he e o e, does no ai h ully ques ion he phenomena ha occu a he le el o he human being. Thus, he use o hese d ugs can lead o complica ions o humans as well as damage o he pha maceu ical indus y. O gan-on-a-chip (OOC) echnology has eme ged wi h he aim o eplica ing impo an aspec s o human physiology and hus being able o o e come he limi a ions o adi ional p ocedu es, he eby inc easing he sa e y and e icacy o hose who ake hem. Al hough OOC echnology p o ides se e al ad an ages compa ed o con en ional echniques, i s ill lacks mul ipa ame e moni o ing sys ems capable o p o iding mic oscale in o ma ion du ing cell cul u e es ing as well as es ing o new d ugs. In his sense, i is e y impo an o de elop mic osenso s in eg a ed in mic o luidic sys ems o moni o he a ious cellula pa ame e s in o de o unde s and wha e ec s he d ugs ha e on cells. Pa ame e s such as empe a u e, oxygen, pH, numbe o nu ien s, among o he s, a e a iables o in e es o he o e all pe cep ion o me abolism c ea ed by he combina ion o d ugs wi h mul iple o gan cells. Thus, he wo k de eloped his disse a ion aims o s udy and de elop he elec onic eading and ac ua ion o a mic o empe a u e senso , based on esis ance empe a u e de ec o (RTD), which will be in eg a ed in o OOC echnology. The sys em de eloped in addi ion o p o iding po abili y and low cos , allows o pe o m eading o se e al mic o senso s wi h a sensi i i y o 15mV/0.1°C. A g aphical componen has also been implemen ed ha allows he use o selec he senso and ack he esul s in eal ime. Keywo ds: Tempe a u e mic osenso s, O gan-on-a-chip, esis ance empe a u e de ec o , mic o luidic sys ems. 15 1. INTRODUCTION In his chap e , i will be made a con ex ualiza ion o he subjec , ollowed by he mo i a ions and objec i es o his disse a ion. When i comes o con ex ualiza ion, i aims a gi ing a li le unde s anding o o gan-on-a-chip (OOC) echnology and he need o senso s o moni o hem. In he ollow-up, he mo i a ion o he de elopmen o his p ojec will be p esen ed, ealizing he eason o choosing his heme. The objec i es o his disse a ion will be p esen ed and desc ibed in de ail, suppo ed by he expec ed esul s o each objec i e. Finally, he s uc u e o he disse a ion will be p esen ed. 1.1 Con ex ualiza ion Acco ding o a s udy by he Ins i u e o Bioma e ials and Biomedical Enginee ing a he Uni e si y o To on o, 20% o enal diseases acqui ed a e hospi al admissions a e di ec ly ela ed o oxici y om e ec s no o eseen in clinical examina ions o hese d ugs. To unde s and he e ec s o a ious d ugs and inc ease he sa e y and e icacy o hose who use hem, OOC echnology is eme ging o simula e impo an aspec s o human physiology when new d ugs a e being de eloped. This echnology allows ec ea ing h ee undamen al aspec s o human physiology: (1) he mul icellula o epi helial ascula in e aces o o gans, in he case o a blood essel, lung o bowel ne wo ks, which unc ion as ba ie s in issues; (2) he o ganiza ion a he le el o he issue o pa enchymal cells, such as li e , hea , muscle skele on o umou s, which a e esponsible o impo an unc ions o he o gan; and (3) sys ema ic in e ac ion be ween o gans, allowing o, mo e accu a ely, pe cei e all he mechanisms in ol ed in he human physiology [1,2]. OOC echnology no only inco po a es a ious ypes o cells, bu also enginee ing aspec s, such as he inco po a ion o di e en ypes o senso s and mic o luidic channels [3]. Tempe a u e measu emen s a e ex emely necessa y o mos biological and chemical p ocesses. This measu emen mus be e ec i e and accu a e, howe e , when he samples a e in mic o o e en nano o de , his ask becomes qui e di icul [4]. RTDs a e highly linea and sensi i e senso s, wi h a e y wide empe a u e ange be ween -200°C and 850°C, always depending on he ma e ial om which i is ab ica ed, and a e he e o e mos ly used o 16 minia u ized empe a u e measu emen sys ems especially due o hei educed dimensions [4,5]. This disse a ion will explo e solu ions o empe a u e mic osenso s in eg a ed in o mic o luidic pla o ms o in i o es ing ocusing on he OOC echnology, in i s his o y and wo k al eady de eloped. The choice o RTD as a empe a u e senso o his wo k compa ed o o he senso s o he same pu pose, i s a chi ec u e, ab ica ion, measu emen echniques, and i s ole in OOC echnology will also be analysed. 1.2 Mo i a ion and obje i es Science and echnology ha accu a ely s udies and manipula es mic oscale luids a e called mic o luidic, and OOC is a subca ego y o hese whose main objec i e is o simula e he physiological aspec s o he human o gan. OOC echnology was conside ed one o he op 10 eme ging echnologies o 2016 a he Wo ld Economic Fo um [6]. Since his is an eme ging echnology, al hough al eady widely used and de eloped, is s ill in an emb yonic s a e and so he e a e s ill some sho comings in i s implemen a ion. One o he s ands o his echnology ha s ill lacks wo k is he enginee ing a ound echnology, in a mo e conc e e case, he in eg a ed senso s. This is a echnology ha o be unc ional equi es senso s ha allow use s o moni o he en i onmen al pa ame e s o he cell, such as pH, oxygen le els, nu ien le els, empe a u e, among o he s. The main mo i a ion o his disse a ion is o s udy and de elop a sys em o con ol and eading o empe a u e mic osenso s in eg a ed on OOC o o e come he ac ual limi a ions. These mic osenso s we e de eloped in he labo a o y o Mic o/Nano echnologies and Biomedical Applica ions o he Cen e o Mic oEle oMechanical Sys ems (CMEMS) o Minho Uni e si y. The main objec i e is he de elopmen o elec onic ci cui s o inc ease he eading esolu ion o hese senso s, using low noise ci cui s and po able sys ems. This sys em mus be able o p ecisely and accu a ely ead wi h a esolu ion o 0.1°C in he ange o 35°C o 45° o co e ing he op imal empe a u e o he human body. Besides, i is also aimed o de elop a g aphical applica ion ha allows he use o iew he empe a u e da a in eal- ime and change be ween he a ious senso s ha can be connec ed o he ci cui . To accomplish his goal, some asks mus be done: 17 • De elopmen o a cu en sou ce o RTD exci a ion The RTD is a empe a u e senso ha uses an exci a ion cu en , known and cons an , o p oduce a ol age in i s e minals, esul an om he combina ion o he known cu en and i s a iable esis ance alue a he poin o measu emen (Ohm law). This cu en alue mus be he maximum cu en ha gi es he bes esolu ion and a he same ime p e en s he senso om sel -hea ing. • De elopmen o low noise and po able con ol and eadou ci cui To p o ide low noise, he powe sou ce o he ci cui mus be independen o he powe g id and his is only possible wi h ba e ies, allowing no only low noise bu po abili y o he sys em. The ci cui mus be in eg a ed in o a PCB o con ibu e also o educing he noise om wi es and he elec onic componen s ha compose i , all mus be low noise. • De elopmen o a g aphical in e ace To see he changes in empe a u e and esul s o mul iple senso s, i is necessa y o ha e an use - iendly in e ace ha can display he alues o empe a u e om an RTD ha is connec ed o he sys em. Also, he use mus be capable o selec ing which senso wan s o ead and sa e he alues on a ile. 1.3 Disse a ion s uc u e This disse a ion is s uc u ed in he ollowing way: • Chap e 1 con ains he in oduc ion, a sho con ex ualiza ion and he mo i a ions and objec i es o his disse a ion. • Chap e 2 p esen he s a e o a , highligh ing he esea ch abou heo e ical subjec s and wo k done in mic o luidic de ices, OOC and empe a u e senso s. • Chap e 3 p esen s he adop ed me hodology o sys em de elopmen . • Chap e 4 con ains he implemen a ion o he de eloped sys em, which ollows he me hodology ou lined in he p e ious chap e . • Chap e 5 add esses all he expe imen al esul s ob ained in he es s made as well as i s analysis and discussion. Some imp o emen s o he de eloped sys em a e also p esen ed. 18 • Chap e 6 p esen s he conclusions d awn om he de elopmen and alida ion o he sys em. Mo eo e , his chap e p oposes u u e wo k o imp o e he sys em. 19 2. STATE OF ART In his chap e , he heo e ical componen o his disse a ion will be p esen ed. Fi s ly, he mic o luidic de ices will be add essed, ollowed by a deepe ocus on he OOC. A e hese opics i will be in oduced he mos common empe a u e senso s used o hese applica ions, dedica ing a sub-chap e o he RTD, he senso ha was chosen o his p ojec . 2.1 Mic o luidic sys ems Mic o luidics is he science and echnology ha p ocess o manipula e small amoun s o luids, using channels wi h dimensions o ens o hund eds o mic ome es. These sys ems in ol e se e al ields: biode ence, mic o echnology, molecula analysis, and molecula biology, among o he s [8]. 2.1.1 His o y A lo o di icul ies and p oblems in a ious ields con ibu ed o he de elopmen and g ow h o mic o luidics. A e he end o he cold wa , mic o luidic sys ems we e impo an o de ec chemical and biological h ea s since his h ea was a majo conce n in his e a. Then, ano he ac o was he big necessi y o mic oanalysis, in he 1980s, in he ield o molecula biology, mainly on high- h oughpu DNA sequencing. This p ocess came wi h he neediness o sys ems wi h mus highe h oughpu , highe sensi i i y, and esolu ion. Mic o luidics o e came hese p oblems. A las , ano he majo con ibu ion o he de elopmen and g ow h o mic o luidics came wi h mic oelec onics. Pho oli hog aphy and associa ed echnologies a e success ully implemen ed in silicon mic oelec onics and MEMS ab ica ion. The objec i e was o ex end his ab ica ion o mic o luidics. This ype o ab ica ion con ibu ed di ec ly o he educ ion o he size o hese sys ems. Howe e , silicon and glass used in he i s wo ks we e apidly eplaced by polyme s because hey a e expensi e and in addi ion o ha he ab ica ion o he mic ochannels is mo e complex and expensi e. Also, he silicon is opaque which b ings p oblems o op ical applica ions. Thus, mic o luidic apidly s a ed o be ab ica ed in polyme ic ma e ials like polydime hylsiloxane (PDMS) [8]. 20 2.1.2 The impo ance o mic o luidic sys ems Mic o luidic sys ems a e low-cos and minia u ized de ices wi h a sho eac ion ime and high h oughpu using a small olume o eagen s/samples which allows he de ice o ope a e nea i s wo king en i onmen . Fu he mo e, hese de ices a e ab ica ed by easy and well-es ablished p ocesses. Ope a ing close o he en i onmen allows high pe o mance as hey a e highly accu a e and e ec i e [9]. These ad an ages ha e eased he selec i e and e ec i e es ing o chemicals in cellula en i onmen s, he e o e addi ionally pe mi ing o obse e in eal- ime he changes ha his has a he physiological le el o he cell, which was p e iously p ac ically impossible. I is a echnology wi h a lo o po en ials ha may gi e dis inc i e oppo uni ies, o ins ance, o a e ci cula ing umou cell isola ion and de ec ion om he blood o pa ien s, which u he s he esea ch o cance s em cell bioma ke s and expands he unde s anding o he biology o me as asis [10]. 2.1.3 Physics o mic o luidics Di e en p ope ies a e shown in he luid low o mic o luidics [9]: Lamina low The eloci y a luid s eam is called lamina low. Due o he small size o mic ochannels, he low is almos always lamina . Lamina low can be desc ibed by Reynolds numbe ,𝑅𝑒, which is de ined by he equa ion, 𝑅𝑒 = 𝜌𝑣𝐷ℎ 𝑢 whe e 𝜌 is he densi y o he low, 𝑣 is he low eloci y, 𝐷ℎ is he hyd aulic diame e o he mic ochannel and 𝑢 is he iscosi y o he low. Di usion Di usion is esponsible o mixing he di e en luid lows ha a e in con ac bu ne e mix due o he muddy low. This can be desc ibed by he equa ion, 𝑡 = 𝑙2 2𝐷 21 whe e 𝑡 is he di usion ime o he molecule o di use o e he dis ance 𝑙 and 𝐷 is he di usion coe icien o he molecule in solu ion. Capilla y numbe This mic o luidics physics is exclusi e o d ople mic o luidics. The con inuous phase low ha de o ms he d ople is de ined by he capilla y numbe , 𝐶𝑎. I is de ined by he ollowing equa ion, 𝐶𝑎 = 𝜂𝑐 𝑣𝑐 𝛶 whe e 𝜂𝑐 is he iscosi y o he con inuous phase, 𝑣𝑐 is he eloci y o he con inuous phase, and 𝛶 is he in e acial ension be ween he oil and wa e phases. Webe numbe Webe numbe is he esul o mul iplying he Reynolds numbe and he capilla y numbe and cha ac e izes he in e acial ension, equen ly used o pa ame ize d ople b eakup p ocesses. 𝑊𝑒 = 𝑅𝑒 × 𝐶𝑎 Bond numbe Bond numbe cha ac e izes he ela i e impo ance o g a i y and in e acial ension. I 's de ined by he equa ion, Bo = ∆𝜌𝑔𝑤2 𝛶 whe e ∆𝜌 is he densi y di e ence be ween he wo immiscible luids, 𝑔 is he accele a ion o g a i y, w is he cha ac e is ic leng h scale, and 𝛶 is he in e acial ension be ween he oil and wa e phases. 2.1.4 Types o mic o luidics Mic o luidics can be di ided in o wo ypes [9]: Con inuous- low mic o luidics This ype o mic o luidics is based on a single-phase luid in he chip. This has he objec i e o educing he consump ion o eagen s and samples, which leads o g ea e speed and ep oduc ion compa ed o adi ional expe imen s. Whoe e , his ype has some disad an ages, like (1) he was e o luid de i a ed om he o e low in he channel, (2) he 22 slowness in achie ing uni o m mixing due o he lamina low, (3) he channel easily con amina es he luid and he eagen , (4) he size o he de ices inc eases wi h he inc easing o he numbe o pa allel expe imen s. This echnique can be seen in igu e 1 [11]. D ople mic o luidics D ople mic o luidics is based on, a he han ha ing a cons an low o luid, only disc e e d ople s a e eleased using immiscible phases. The eac ion o each d ople can be handled wi hou he in e e ence o eagen s o he sample. Also, his ype o mic o luidics has a highe h oughpu and p o ides a pla o m o pe o m many eac ions. The e a e h ee main echniques used in his sys em: (1) co- lowing, (2) T-junc ion and (3) low- ocusing ( igu e 2) [9]. Figu e 1 – Schema ic ep esen a ion o a con inuous low. Adap ed om [11]. (1) Capilla y (2) T-Junc ion (3) Flow ocusing Aqueous Oil Aqueous Oil Aqueous Oil Figu e 2 - Types o mic o luidics echniques o d ople s gene a ion. Adap ed om [8]. 23 2.1.5 Ma e ials o mic o luidic de ices The mic o luidic de ices ha e been widely ab ica ed by PDMS-based so li hog aphy since his ma e ial is easily ab ica ed and easily bond h ough plasma ea men . Fu he mo e, he PDMS is a low-cos ma e ial, biocompa ible wi h high op ical anspa ency and has good gas pe meabili y and high he mal s abili y. I also eplica es nanome ic cha ac e is ics om moulds ab ica ed by UV pho oli hog aphy, which can be eused se e al imes, ensu ing ep oducibili y. As i has supe icial ea men me hods well de ined enables he unc ionaliza ion o he de elopmen o biosenso s [12]. Howe e , he moulding o hese ma e ials in ol es a la gely manual p ocess, which makes he manu ac u ing p ocess di icul o au oma e and consequen ly expensi e [13]. The moplas ics, like polys y ene, PMMA, polyu e hane, among o he s, a e used o ab ica e mic o luidic de ices, enabling high- h oughpu bu compa ed o he PDMS, does no allow ab ica ion wi h high-quali y. This ype o ma e ial will bend a ound 1000 imes less han he PDMS because i s Young’s Modulus is h ee imes la ge han he PDMS, which in some applica ions can be an ad an age [14]. Finally, in 2008, a s udy p oposed c ea ing de ices in pape o sol e he p oblems o cos s and he equi emen o pumps. Pape is an inexpensi e ma e ial ha can pump he luid i sel wi hou he need o pumps, by is wicking ac ion, and i is accessible o anyone ha has a simple wax p in e . Howe e , i has i s disad an ages, namely, i is no a good ma e ial o cell manipula ion because i abso bs biomolecules which can di icul he de ec ion o some analy es, especially when hey a e in lowe concen a ions. Fo now, i is no an ideal solu ion o be used in bio echnology [10]. In conclusion, he e is a need o imp o e he ab ica ion p ocesses o educe he o al cos s. I is needed o de ices o be made o a ma e ial ha can be apidly p o o yped, like PDMS o pape , ab ica ed o a low cos pe uni , and whe e he low can be ully con olled, s opped, and e-s a ed. 24 2.2 O gan-on-a-chip OOC is a mic o luidic de ice specialized in mimicking he unc ions o human o gans in i o. I can ec ea e he en i e cellula en i onmen , he connec ions be ween cells and hei in e ac ions. I allows o p ecisely egula e he condi ions whe e cells a e g owing and adap hei ope a ion wi hin he en i onmen ha is being ec ea ed. Thus, egula es all pa ame e s and op imizes he cell cul u e mic oen i onmen [16]. P elimina y es ing o d ugs be o e hey a e a ailable o humans is he mos impo an p ocess, bo h on he consume side, o ensu e ha hey do no su e om se ious side e ec s, as well as on he pha maceu ical indus ies side, as hese es s incu high cos s and i is impe a i e o assu e he sa e y o he consume . The de elopmen o new d ugs goes h ough a s age o animal es ing, which some imes p oduces a high ailu e a e because hey do no eplica e human physiology e y accu a ely. In addi ion, hey a e ime-consuming p ocesses and ha e e hical issues ha g ea ly limi hei use. So, he es in animals has been complemen ed by 2D cell cul u es in i o [1,3]. Howe e , in i o models, e eal one majo disad an age compa ed o he in i o me hods, hey canno , mos o he ime simula e he complex cell-cell and cell-ma ix in e ac ions c ucial o egula ing cell beha iou in i o. Despi e ha , hey a e use ul o s udying he molecula basis o physiological and pa hological esponses. The p oblems wi h es ing in 2D models o animal-based models ha e gene a ed sizeable in e es in he sea ch o human- based issue-like cons uc s o imp o ing he disease modelling and d ug and chemical es ing, inc easing hei e ec i eness and educing economic losses om he lack o e ec i eness [17]. Since he 90s, se e al mic o luidic echnologies ha e been applied o in i o sys ems. Mic o echnology has opened doo s o OOC echnology o se e al easons: he ac ha makes i possible o implemen small elec ic and luidic ci cui s o be in eg a ed in o minia u ized de ices. They also allow g ea e p ecision and acili a e he deli e y o essen ial nu ien s o cell g ow h, due o hei lamina na u e, so he g adien s a e o high concen a ion due o a sho e di usion leng h. Finally, as he las ad an age, he shea s ess applied o he cells can be con olled and minimized. In a way, all hese ad an ages allow he c ea ion o a well-con olled en i onmen in he in i o cul u es wi hin OOCs, and hus 31 hand ha e a nonlinea esponse, small sensi i i y, small ou pu ol age, equi e cold junc ion compensa ion and a e leas s able. The he mis o is a as esponse senso , has a high ou pu and minimal lead esis ance bu needs an ex e nal powe o ol age sou ce, is non-linea , e y agile and su e s om sel -hea ing e o s. The RTD is he mos s able wi h good linea i y and accu acy (0,01°C) bu has low sensi i i y, a ound 0,00385 Ω/Ω/°C o pla inum senso s, i is also ex e nally powe ed wi h a cu en sou ce, i is mo e expensi e, has a small ou pu esis ance and also su e s om sel -hea ing e o s. Howe e , he RTD a e easie o minia u ize and ab ica e in o he mic o luidic de ices and wi h sui able ampli ica ion and il a ion elec onic ci cui s, a high sensi i i y esponse can be ob ained [31,36]. In he de elopmen o a senso , i is impo an o ake in o accoun ce ain cha ac e is ics app op ia e o he desi ed applica ion [32]: Ope a ion in e al In he ope a ion in e al, he empe a u e limi s ha he senso eads, lowe and uppe , a e de ined. A longe in e al can some imes p o ide a lowe senso 's linea i y and sensi i i y. Linea i y Is he capabili y o he senso o espond linea ly o all he changes o empe a u e h ough all he ope a ion ange. Sensi i i y I is said ha a senso is mo e o less sensi i e depending on he change in i s ou pu alue when he empe a u e changes, ha is, wi h a small empe a u e change, he mo e a iabili y he e is in he ou pu alue, he mo e sensi i e i is he senso . Response ime When a empe a u e change occu s, he ime i akes o he senso ou pu o change i s alue is gi en wi h he esponse ime o he senso . S abili y This cha ac e is ic is de ined by he senso 's abili y o main ain he same esul a he ou pu when i is unde a s able empe a u e. 32 E icacy The e ec i eness o he senso is gi en by he en i onmen ha su ounds i and he ci cui s whe e i is implemen ed. To be mo e e ec i e i is no mally needed good wi ing, good g ounding h oughou all ci cui s, noise il a ion, among o he s. Du abili y The du abili y o he senso is he capabili y o he senso o ope a e in he same en i onmen o some ime, no mally decided by he manu ac u e . This cha ac e is ic is usually ela ed o he ma e ial ha composes he senso and i s encapsula ion. Requi emen s o condi ioning signal All senso s will ha e a ious signal condi ioning equi emen s such as ampli ica ion, cold-junc ion compensa ion ( he mocouples), il e s, exci a ion (RTD and he mis o s), o se e o adjus men , scaling o empe a u e uni s, lead esis ance co ec ion, channel- o-channel isola ion, open he mocouple de ec ion ( he mocouples only). Cos s In all p ojec s, i is impo an o deno e he cos s, some imes a cheape senso will sa is y he needs o he p ojec , as i will p obably need less su ounding ma e ial like wi ing, signal condi ioning o e en assembly cos s. Howe e , some applica ions need mo e complex senso s and hus op imized manu ac u ing, which can inc ease he inal cos . Ne e heless, he le el o in eg a ion o he sensi i i y equi ed may jus i y a highe cos . 2.4 Tempe a u e measu emen in mic o luidic de ices Se e al wo ks abou empe a u e measu emen s in mic o luidic sys ems ha e been s udied and implemen ed. In he i s example, i is p esen ed a nanoneedle o measu e he empe a u e o single cells. This sys em s udies he possibili y o a low o wa e o mo e he cells agains he nanoneedle wi h enough amoun o o ce o pe mi he nanoneedle o pene a e he cell wi hou damage. An elec ical ol age is applied o he nanoneedle and i is measu ed he ou pu cu en o he con e sion o empe a u e, howe e , he s udy was only simula ed and no implemen ed, he esul s we e p omising [38]. I is inc easingly impo an o analyse he beha iou o cells, bu i is also necessa y o be ca e ul o no damage hem. The e a e cells om diseases such as cance , which e eal an 33 inc ease in empe a u e which makes i impo an o s udy in e cellula empe a u e. To measu e cell empe a u e, many a emp s we e made, success ully and wi h high le els o sensi i i y, bu he ma e ials used in hese measu emen sys ems p o ed o be a c i ical disad an age because o hei high oxici y wi hin he cell in luencing hei de elopmen . These phenomena occu ed when empe a u e sensi i e-ma e ials we e used o measu e he empe a u e o cells, such as quan um do s, ha a e used as nano he mome e s [39], nanopa icles, ha a e connec ed wi h o he nanopa icles and oge he ac as a molecula sp ing and wi h he a ia ion o he empe a u e i s ex ension changes and ha ex ension is used o measu e he empe a u e [40], and he mosensi i e ma e ials [41]. Nanoneedles ollowed, accompanied by ma e ial wi h p ope ies ha a e a ec ed by empe a u e change. In he s udy o Binsilm e al. [38], a nanoneedle mic o luidic sys em was p esen ed. The based ma e ial used in i s ab ica ion was he ungs en and he senso showed o ha e a posi i e empe a u e coe icien o empe a u e (TCR) wi h he inc ease o empe a u e. Fab ica ed in Y shape allowed he nanoneedle pene a ion in he cell o cause minimal damage o he cell. The ou pu o his senso is gi en in he cu en o m, and i was ob ained a esolu ion o 0.02°C which ansla es in a ia ions o 50nA. In a s udy by Chunyan Li e al. [42], is p esen ed he de elopmen o a sma mic oca he e wi h se e al in eg a ed mic osenso s, such as a empe a u e senso (RTD), low a e senso (ho - ilm anemome y) and glucose biosenso (ampe ome ic senso ). The empe a u e senso was ab ica ed in i anium gold (Ti/Au) deposi ed in a polyme ic and lexible subs a e. A p ecise and linea esponse conce ning empe a u e was demons a ed using he ab ica ed RTD. Fu he mo e, he au ho s showed ha he ab ica ion p ocess o mic oca he e wi h in eg a ed mic osenso s is e y simple and sui able o low-cos mass p oduc ion. Ano he app oach o empe a u e senso s in he mic o luidic sys em was ela ed by Naoki Inoma a, e al. [43]. This app oach was based on a pico calo ime e ha uses a esona o o de ec he empe a u e a ia ion in he sample by equency acking o he esona o . The sys em has a hea guide ha conduc s he hea o he sample and he esona o is kep in acuum. The change o empe a u e makes he esona o ib a e and ha ib a ion equency was ansla ed in he a ia ion o empe a u e in he mic o luidic channel. Was achie ed a esolu ion o 5.2 pJ in he ab ica ed senso . 34 Mic o luidic de ices a e no only used in medical expe imen s. In he wo ks o Schmi e al. [44] was p oposed he implemen a ion o an in elligen mic o luidic empe a u e indica o o p ecisely p edic he expi ing da e o he ood on he package. The idea goes h ough a sma package wi h passi e senso s ha , based on he empe a u e and ime ha ood is packaged, demons a es a mo e eal expi a ion da e han con en ional ones p in ed a he ime o packaging ha is es ima i e. Senso s ha al eady exis a e usually on he ou side o he package and use ead ci cui s ha do no make i e y easy o ead. This s udy p oposed a senso wi h wi eless eading. The senso is composed o a capilla y wi h a ec angula c oss- sec ion wi h wo elec odes placed in opposi e walls o ob ain a pla ed capaci a o . Non- oxic luid is used wi h high iscosi y ha will mel , educing he iscosi y, wi h he inc ease o empe a u e, en e ing he mic ocapilla y, and inc ease i s iscosi y wi h he dec ease o empe a u e, decele a ing he luid and s opping i . The dis ance ha he luid a elled will be ela ed o he capaci y be ween he wo elec odes, which inc eases p opo ionally wi h he inc ease in empe a u e. Finally, an RTD senso was bonded oge he o he su ace o a PDMS based he moelec ic coole in a wo k o Oile e al. [45] o a he oscle osis s udies. As he p e ious wo ks ha used ho -wi e anemome e s, al hough ha ing high sensi i i y and being obus , had a limi a ion in low p ope ies measu emen a cell le el because i was p o en ha damages he cells. I is impo an o unde s and he blood low in a he oscle osis s udies because his disease a ec s di ec ly he blood essels bi u ca ions and i is de ec ed by s udying he shea s ess ha is di ec ly co ela ed wi h he low a e. The he moelec ic coole was ab ica ed in PDMS and he RTD was bonded o he su ace o he de ice, his allowed each ime ha he low was inc eased, he empe a u e senso de ec ed ha inc ease by changing i s alue o empe a u e. This p o es ha i is possible o use he RTD in a he moelec ically coole o measu e he low a e o blood. 2.5 Tempe a u e mic osenso s based on RTD Measu ing empe a u e wi h con en ional senso s such as he mocouples and he mis o s a e success ul in samples in he o de o millime es, bu when ad ancing o smalle samples hese senso s a e mo e di icul o in eg a e, as an example, in o mic o luidic de ices. Se e al app oaches o cell cul u e empe a u e sensing ha e been explo ed in 35 p e ious wo ks such as comme cial PT-100 RTDs [46], luo escen polyme ic he mome e s [47] and comme cial T- ype he mocouples [48]. Howe e , hese solu ions a e no e y compac o in eg a ion in mic o luidic de ices o p esen a poo esolu ion. The ad ances in mic o echnology ha e allowed he de elopmen o minia u ized empe a u e senso s based on RTDs which can be in eg a ed wi hin mic o luidic de ices, as p oposed in his disse a ion [4]. 2.5.1 Cha ac e is ics To cha ac e ize he RTD's, some de ails abou i should be speci ied [49]: Type o esis i e ma e ial Mos RTD is ab ica ed wi h pla inum, al hough o he ma e ials as nickel, aluminium, coppe can be used. Pla inum is he p e e ed ma e ial since has highe linea i y and sensibili y. Tempe a u e coe icien To con e he senso ou pu alues o empe a u e, i is necessa y o pa ame e ize he empe a u e coe icien which ep esen s he change o esis ance pe uni o he empe a u e o he esis i e ma e ial om 0°C o 100°C ( eezing and boiling empe a u es o he wa e , espec i ely). The empe a u e coe icien decla ed by he IEC (In e na ional Elec o echnical Commission) in 1983 o he pla inum was 0.00385 ohms pe ohm deg ee cen ig ade. This alue is accep ed wo ldwide when using pla inum RTDs. Nominal esis ance This ea u e is he empe a u e ha should be expec ed o a ious empe a u e alues. 100 ohms is he s anda d IEC a 0°C o PT100 bu he e a e o he nominal esis ances a ailable (PT500, PT 1000, and o he s). Tempe a u e ange o ope a ion As i s name implies, i is he empe a u e ange whe e he senso ope a es. This alue a ies wi h i s con igu a ion and ab ica ion me hods. Dimensions and size es ic ions The esis i e elemen mus i inside i s p o ec i e encapsula ion, so i s mos impo an dimension is he ou e diame e . In his case, he RTD a e easy o minia u ize acili a ing i s in eg a ion. 36 P ecision This speci ica ion is gi en acco ding o he IEC 751 o pla inum RTD's as we can see in able 1, di ided by wo classes (class A and class B) ollowing DIN 43760 equi emen s o accu acy. Response ime I is he ime ha he senso akes o espond o a change o empe a u e and p o ide a s able alue. Sel -hea ing This is a pa ame e ha is e y impo an o be awa e o because i ansla es in pe u ba ions on he alues on he eading. As hese senso s need a cu en o exci a ion o ope a e, ha alue mus be chosen and pa ame e ized so he senso ne e en e s in sel - hea ing. This e o is ansla ed in o °C/mW. Wi e con igu a ion The elec ical connec ions o he exci a ion and eading o he RTD senso s add elec ical esis ance ha is summed up o he esis ance o he senso . In his way, when Tempe a u e °C Class A Class B De ia ion De ia ion Ohms °C Ohms °C -200 ±0.24 ±0.55 ±0.56 ±1.3 -100 ±0.14 ±0.35 ±0.32 ±0.80 0 ±0.06 ±0.15 ±0.12 ±0.30 100 ±0.13 ±0.35 ±0.30 ±0.80 200 ±0.20 ±0.55 ±0.48 ±1.30 300 ±0.27 ±0.75 ±0.64 ±1.80 400 ±0.33 ±0.95 ±0.79 ±2.30 500 ±0.38 ±1.15 ±0.93 ±2.80 600 ±0.43 ±1.35 ±1.06 ±3.30 650 ±0.46 ±1.45 ±1.13 ±3.6 Table 1 - Pe missible de ia ions om heo e ical pla inum RTD alues. Adap ed om [34]. 37 accu acy is impo an , he wi e con igu a ion needs o be conside ed o minimize ha in luence. 2.5.2 Fab ica ion The e a e RTD's ab ica ed wi h conduc i e wi e w apped in a glass o ce amic co e, howe e , he minia u ized RTD's a e usually based on esis i e hin- ilm o pla inum, nickel, o aluminium. Pla inum RTD's a e he mos accu a e because hey demons a e g ea e s abili y o e ime. Tempe a u es can be measu ed in a ange o -189°C o +962°C wi h a maximum accu acy o 0.001°C [36]. The hin p o ile o he RTD allows i o be ab ica ed in a plane wall wi h a hickness o ew nanome e s and hus, can be easily in eg a ed in o mic o luidic de ices. Usually, glass o PDMS a e used o close he mic o luidic channels and RTD can be di ec ly ab ica ed on o his subs a e. The RTD is ab ica ed by hin ilm acuum deposi ion echniques and hei s anda diza ion is ca ied ou by pho oli hog aphy echniques and li -o p ocess [44,45]. Fi s , i is necessa y o ealize ha he ab ica ion o nanoma e ials and nanos uc u es is pe o med in an ex emely clean en i onmen called a clean oom, in which he e a e ew pa icles and specialized equipmen o h ee basic unc ions: (1) Thin ilm deposi ion; (2) Pa e ning; (3) E ching. The hin ilm deposi ion equi es a acuum en i onmen in he o de o 10−6 o o 1.3 × 10−4 Pa o ensu e ha any impu i ies o con aminan s a e no p esen in he deposi ion chambe . The ma e ials o be deposi ed mus be o high pu i y, ypically 99.9999% o highe . The h ee mos common hin ilm acuum deposi ion echniques a e: The mal e apo a ion The mal e apo a ion uses hea o e apo a e he sou ce ma e ial o high pu i y om solid ma e ial o gas. The gas a oms a el h ough he acuum inside he chambe and when hese a oms each he sub ac , hey condense and o m a hin ilm on he su ace. Me als and dielec ics can be deposi ed using his echnique bu a e mos commonly used o deposi me als because hei boiling poin is much lowe and hei deposi ion a es a e mo e s able. This p ocess is comple ely au oma ized, a compu e con ols he empe a u e inside he chambe and he hickness o he hin ilm, i jus needs he use o inse a he beginning o he p ocess he hickness desi ed and he limi o empe a u e [52]. 38 Elec on beam e apo a ion Elec on beam e apo a ion is simila o he mal e apo a ion. I also uses hea o make he e apo a ion o he high pu i y ma e ial and is pe o med inside a acuum chambe . The majo di e ence is he sou ce o hea . A beam o elec ons is esponsible o hea ing and consequen e apo a ion o he ma e ial o be deposi ed [53]. The elec on beam allows he hea ing o a me al ilamen un il abou 2500°C. A his empe a u e, he elec ons a e e y ene ge ic and some o hem lea e he su ace o he ilamen and a e accele a ed in di ec ion o he sou ce ma e ial wi h he help o a high ol age elec ode. A g oup o magne s lead he elec on and ocus he beam o he sou ce ma e ial [53]. One o he ad an ages o his echnique is ha i allows being deposi ed di e en ma e ials on he subs a e wi hou opening he chambe because he elec on beam is well con ined in space, i hea s only a small a ea o he sou ce ma e ial allowing o he sou ce ma e ials o be in he chambe o be e apo a ed sequen ially. The sou ce ma e ial is pu inside he chambe on he c ucible [53]. Sys ems ha hold ou ma e ials a e e y common and hey a e called he ou -pocke hea h. Fou c ucibles i in o he hea h, and each c ucible can hold a di e en sou ce ma e ial. So ha i can ha e up o ou laye s o di e en ma e ials deposi ed wi hou b eaking he acuum. The hea h is a o a ed holde o coppe which is wa e -cooled. The wa e cooling p e en s he c ucible ma e ial om mel ing and mixing wi h he sou ce ma e ial o wi h he hea h i sel [53]. Spu e ing deposi ion Spu e ing is a di e en p ocess compa ed o he o he wo ypes because i does no use e apo a ion. Ins ead, uses ene gized a oms ha each he sou ce ma e ial ( a ge ) o be deposi ed on he subs a e. These ene ge ic a oms hi he a ge and p opel he a oms o he sou ce ma e ial ou o he a ge and in o he acuum sys em. Some o hese a oms each he subs a e ha is a anged on he op o he acuum chambe and deposi . These ene ge ic a oms used in he p ocess a e c ea ed by plasma. The a gon is he gas used o o m plasma because i is an ine gas and will no esul in unwan ed chemical eac ions du ing he p ocess [54]. 39 When he gas is in he chambe , a high elec ical ol age is applied o he a ge . This high ol age is s ong enough o emo e an elec on om he a gon a oms causing i o become ionized and as he sou ce ma e ial is nega i ely cha ged, he a gon a oms will eac wi h hese a oms indi idually by emo ing hem om he a ge ma e ial and peppe ing hem in a ious di ec ions, including owa d he subs a e [54]. Pho oli hog aphy Pho oli hog aphy is an indispensable s ep in he ab ica ion o hin ilm-based RTD. This echnique is esponsible o ans e he desi ed pa e n o he senso o a pho o esis ma e ial which is used o p o ec selec ed a eas du ing subsequen e ching o deposi ion p ocesses. I uses ul a iole ligh o ans e he geome ic pa e n om an op ical mask o a pho o esis deposi ed on he subs a e [55]. 40 3. METHODOLOGY In his chap e is p esen ed he me hods and he alida ion o a empe a u e senso based on RTD. All he p ocedu es inhe en o he de elopmen o hese sys ems a e de ined. The gene al equi emen s will be po ayed in he ollowing. • Reading esolu ion o 0.1°C, in he ange o 35°C o 45°C; • Low noise ci cui ; • Real- ime and in si u moni o ing; • Display he collec ed da a in g aphical o m; • Compac size; • Po able sys em; • Capabili y o ead mo e han one senso . The main goal o his disse a ion is he de elopmen o a sys em o measu e empe a u e changes in OOC wi h a esolu ion o 0.1°C in a ange o 35°C and 45° C. This ange o empe a u e alues co e s he ypical human body empe a u e, a ound 37°C, whe e he human cells ypically g ow and i also can be used in s udies o hypo (35°C) o hype he mia (42°C). To achie e his goal, he noise in he ci cui mus be minimum because in low signal sys ems like his, e e y pe u ba ion, e en i i is minimal, is going o ha e a big impac on he signal and, consequen ly on he alues ha a e ead and con e ed. A las , he alues ha a e ead and con e ed a e going o be displayed in g aphical o m o be easy o he use o ollow he empe a u e ansi ions o he en i onmen whe e he senso s a e inse ed. A esolu ion o 0.1°C is c ucial o de ec ing, as an example, an exo he mic eac ion esul ing om he in e ac ion o he cells wi h d ugs. Fu he mo e, he sys em mus be po able and compac wi h he capabili y o eading mul iple senso s. The po abili y and he educed size will allow he sys em o ope a e inside o equipmen wi h small dimensions, like incuba o s whe e he le els o oxygen and ca bon dioxide a e con olled. The RTD senso s will be in eg a ed in o mic o luidic de ices, and hus, he de eloped sys em mus be able o pe o m au oma ic, eal- ime and in si u empe a u e measu emen s. The con ol and he eading sys em de eloped in his disse a ion was di ided in o di e en pa s o a be e unde s anding o he sys em. Fu he mo e, i simpli ied oubleshoo ing i some hing is no wo king o needs o be imp o ed/changed. 47 In [58] an ins alla ion o a senso using he ou -wi e con igu a ion o geo he mal s udies was add essed. Since i is an applica ion whe e he senso is ins alled a abou 100m dep h, his makes i impossible o he senso ou pu alues o be ead on si e. I is necessa y o use conduc i e wi es o he alues o be ead emo ely. These wi es will ha e a esis ance ha is associa ed wi h hem wha causes he esul s o be changed and a e no eal, so he ou -wi e con igu a ion o his applica ion was chosen. As expec ed, i was concluded ha he e we e no luc ua ions o measu emen e o s de i ed om he esis ance o he lead wi es and all he alues ead on he senso we e he eal ones. 3.1.2 Fab ica ed RTD The empe a u e mic osenso s used in his disse a ion we e de eloped in p e ious wo ks o some esea che s in he labo a o y o Mic o/Nano echnologies and Biomedical Applica ions o he Cen e o Mic oElec oMechanical Sys ems (CMEMS) in pa ne ship wi h he In e na ional Ibe ian Nano echnology Labo a o y (INL) whe e hey we e ab ica ed. Mo e de ail abou he ab ica ion p ocess and cha ac e iza ion o he empe a u e RTD mic osenso s a e desc ibed in [59]. A e some compa ison s udies, pla inum was selec ed o ab ica e he RTD, since i was he one ha ga e be e esul s in e ms o sensi i i y. Fu he mo e, he ab ica ion p ocess is well es ablished p o iding good ep oducibili y. The pa e ning o he senso s was made ia li -o p ocesses and pho oli hog aphy echniques, wi h a se pen ine geome y ha ga e he maximum leng h in a educed a ea allowing he senso o ha e mo e elec ical esis ance and consequen ly mo e sensi i i y. The geome y o he RTD used consis s o meande -shaped wi h 24 windings o a 200 nm pla inum hin ilm (deposi ed by spu e ing, as desc ibed in sec ion 2.5.2) wi h a line wid h and spacing o 10 µm. These senso s we e designed in a ou - wi e con igu a ion, o ensu e high accu acy and an ou pu ol age di ec ly p opo ional o he RTD esis ance only. 48 In igu e 10, i can be seen a wa e wi h a ious and in igu e 11 senso s de ailed pho og aphy o one ab ica ed senso . These senso s ope a e wi h an exci a ion cu en and i was impe a i e o know how he senso esponds o e e y alue o cu en o ind he maximum alue o cu en ha he senso can ope a e wi hou sel -hea ing. P e ious wo ks [59] showed ha 100µA p o ides a negligible sel -hea ing (0.008 °C), and hus, was selec ed o he ollowing es s. In he inal pa o he p ojec , he de eloped sys em will be in eg a ed in o a mic o luidic de ice wi h ou empe a u e mic osenso s and a mic o hea ing sys em. The main objec i e o his in eg a ed sys em is o p o ide empe a u e op imal condi ions o cul u e cells g ow h in OOC. The in o ma ion collec ed by he senso s will be c ucial o ac ua e he hea ing sys em and keep he en i onmen a a empe a u e ha he use wan s. Figu e 10 – Pho og aphy o he ab ica ed RTD. Figu e 11 – Wa e wi h a ious aluminium senso s wi h he same geome y bu di e en windings. 49 3.1.3 Calib a ion echniques To achie e high senso e ec i eness, se e al calib a ions a e equi ed. Depending on he e ec i eness equi ed, calib a ion me hods may be mo e expensi e, i.e., he cos o calib a ion ope a ion inc eases as senso e ec i eness inc eases. Typically, a he indus y le el whe e RTD is used, he empe a u e anges used a e small and his makes i s calib a ion simple [36]. The highe accu acy o a he mome e he mo e expensi e he calib a ion echnique is demanded. Howe e , o some indus ial-g ade RTDs, i is possible o pe o m a simpli ied calib a ion echnique because o he na ow empe a u e anges hey a e used o e . I is also essen ial ha he calib a ion echnique should d aw less ope a o a en ion and op imize calib a ion ime. In he publica ion o V. V. Gu eye e al. [36], a e sa ile semiau oma ic me hod o calib a ing indus ial RTDs was p esen ed wi h he use o he compa ison me hod, compa ing he esponse o he RTD o an indus ial he mome e . The e a e wo calib a ion me hods: Fixed poin calib a ion This calib a ion me hod is e y p ecise, i can each calib a ion close o 0.0001°C and consis s o ixing he senso o be calib a ed inside a chambe wi h a limi ed numbe o ixed hea ing poin s a di e en empe a u es bu each main ains a i s cons an empe a u e du ing he p ocess. These empe a u e alues a e known and selec ed h ough he In e na ional Tempe a u e Scale o 1990(ITS-90) o selec he minimum numbe o poin s. The poin s ha a e in ITS-90 ha e been documen ed and accep ed by he global communi y [60]. Compa ison me hod In his calib a ion me hod, he senso is compa ed o o he s ha al eady ha e been calib a ed and i s esponse is compa ed wi hin an en i onmen and he empe a u e is known and con olled. The e a e se e al ad an ages o using his me hod in compa ison o he ixed poin me hod. Fi s ly, his me hod allows a a ie y o empe a u es, a he han ixed empe a u es and on a limi ed numbe . Ano he ad an age is ha hese calib a ion sys ems a e easy o use and allow calib a ion o mul iple senso s a he same ime and in an au oma ed manne . This me hod is he mos used o calib a e RTD's o indus ial applica ions. Howe e , i is necessa y o pay maximum a en ion and ensu e ha he empe a u e o he he mos a is s able and uni o m o a oid a lack o calib a ion e ec i eness [36]. 50 This me hod was used o he RTD calib a ion using a ho pla e (P äzi he m ype PZ28-2, 1100 W) wi h a esolu ion o 0.1°C which has a come cial senso in eg a ed which i was used o alida ion o he RTD senso used. 51 4. SYSTEM DEVELOPMENT A e he me hodology is de ined, he elec onic ci cui o con ol and ac ua e he RTD will be de eloped. This sys em should be capable o measu ing empe a u e changes, wi h a esolu ion o 0.1°C. 4.1 Con ol & ac ua ion ci cui This is he mos impo an pa o he sys em since a ec di ec ly he pe o mance o he senso s. A sui able cu en sou ce and, ampli ica ion and il e ing ci cui s a e essen ial o ensu e p ope unc ioning. E e y hing has o be well chosen and implemen ed o low noise and accu a e measu emen s. 4.1.1 Powe sou ce The supply o he ci cui was pe o med by ba e ies o assu ing he independence o he powe supply and po abili y. Two ol age egula o s (L7805CV and L7905CV) we e used o supply he ci cui wi h 5V and -5V o powe all he o he componen s o he ci cui - These supply alues we e chosen because he cu en sou ce ope a es wi h 5V, and hus, o be easie o in eg a e and design o he ci cui , all he o he elemen s we e also supplied wi h 5V and - 5V.In igu e 12, can be seen he designed powe ci cui . The capaci o s mus be all elec oly ic and hei alues a e acco ding o he ecommenda ions o he manu ac u e o he egula o s [40,41]. Figu e 12 – Powe sou ce ci cui wi h ol age egula ion and p o ec ions. 52 4.1.2 Cu en sou ce The RTD needs an exci a ion cu en o ope a e and any luc ua ion o he exci a ion cu en will o igina e a change in he elec ical po en ial o he senso s, and consequen ly, will p o ide inaccu a e empe a u e alues. So, an adjus able cu en sou ce (LM134 om Texas Ins umen s, Dallas, TX, USA) was used o p o ide a s able cu en o 100 µA. This de ice was chosen because i allows a cu en adjus men om 1 µA o 10 mA, wi h high accu acy and low noise. In igu e 13, i can be seen he implemen a ion o he cu en sou ce. This implemen a ion is ecommended by he manu ac u e o his ype o ci cui ha is empe a u e sensi i e. The addi ion o a diode (1N457) and a second esis ance (R2) cancels he empe a u e-dependen cha ac e is ic o he LM134 [63]. To p o ide a s able cu en o 100 µA, he alues o R1 and R2 mus be calcula ed using he ollowing equa ions: 𝐼𝑠𝑒𝑡 ≈ 0.134𝑉 𝑅1 𝑅2 = 10𝑅1 Figu e 13 - Cu en sou ce ci cui o powe he RTD. (11) (12) 53 Finally, wi h he alue o 100 µA in Ise , i was ob ained a alue o 1.34kΩ o R1 and 13.4 kΩ o R2, bo h esis ance used a e p ecision esis ances wi h an 1% ola ance. The expe imen al alue o he cu en measu ed wi h he implemen ed ci cui can be seen in igu e 14. 4.1.3 Mul iplexe s To selec he senso o ead, i was chosen he mul iplexe 4:1 (TMUX1104) since his componen allows o ead up o ou senso s. This componen is a low noise p ecision mul iplexe ha has a ail- o- ail logic, ha enables ope a ion in all anges om GND o VDD, ul a-low leakage cu en , ul a-low cha ge injec ion and a ail-sa e logic ha allows he analogue pins o ope a e up o 5.5V ega dless o he ol age in he supply pin [64]. In able 3, he u h able o he componen is p esen ed. Table 3 -T u h able o TMUX1104. Adap ed om [64]. EN A1 A0 Selec ed inpu o D ain(D) Pin 0 𝑋∗ 𝑋∗ All channels a e o 1 0 0 S1 1 0 1 S2 1 1 0 S3 1 1 1 S4 Figu e 14 – Expe imen al exci a ion cu en using he implemen ed cu en sou ce ci cui . *X deno es don’ ca e 54 Fo he selec ion o be done success ully, he sys em mus ha e wo mul iplexe s, one o each g oup o wi es (RTDXWIRE2 and RTDXWIRE3). The signal channels (A0 and A1) o he mul iplexe will be connec ed and oge he connec ed o he wo signal pins o he mic ocon olle , hus ensu ing ha he co ec wi es o he RTD a e chosen simul aneously. 4.1.4 Ampli ica ion Fo he ampli ica ion, i was chosen he high accu acy ins umen a ion ampli ie AD620. Like is impe a i e h ough all he ci cui s is also a low noise componen wi h an excellen DC pe o mance, i has a gain ange om 1 o 10 000 wi h high accu acy and low cos . I 's ideal o his sys em as i has also a low powe consump ion [65]. AD620 has i s gain p og ammed by a esis o (Rg) o o he impedance. The equa ion o calcula e he Rg is he ollowing: 𝑅𝑔 =49,4𝑘Ω 𝐺 − 1 The ou pu ol age o he empe a u e senso s used in his disse a ion shows high linea i y be ween empe a u e and elec ical po en ial, anging om 163.5 mV o 167.5 mV when he empe a u e is inc eased om 35 o 45°C. The ADC has a maximum inpu ol age s a ed a 3.3 V, so, a gain o 16 was de ined, in his phase, allowing he ou pu o he senso o ange om a ound 2.53 V o 2.6 V. Applying he o mula, he alue o he Rg is a ound 3.3kΩ. (13) 55 Finally, he ci cui esponsible o he selec ion and ampli ica ion o he signal can be seen in igu e 15 4.1.5 Fil e ing The ampli ica ion sys em is low noise bu he e was a need o a enua e he signal be o e i en e s he ADC o be e eading. I was implemen ed a i s -o de low pass il e wi h a cu ing equency o a ound 106 kHz. This equency gi es he be e a enua ion o he signal. Wi h he alue o he equency, he alues o he esis ance and he capaci o ha ake pa in he il e can be calcula ed wi h he ollowing o mula: 𝐹𝑐 =1 2𝜋𝑅𝐶 whe e Fc is he cu ing equency (Hz), R he esis ance (Ω) and C he alue o he capaci o (F). The alue o he esis ance calcula ed was 15 Ω and he capaci o was 100 nF. In addi ion o ha , an inpu il e was also implemen ed o p e en adio equency ec i ica ions e o s in ins umen a ion ampli ica ion. This ype o e o appea s as a small dc ol age o se and no mally occu s because all ins umen a ion ampli ie s ec i y small ou o band signals. The inpu signal is limi ed by he il e acco ding o he ollowing wo equa ions [65]: 𝐹𝑖𝑙𝑡𝑒𝑟𝐹𝑟𝑒𝑞𝐷𝐼𝐹𝐹 = 1 2𝜋𝑅(2𝐶𝑑+ 𝐶𝑐) Figu e 15 - Ampli ica ion ci cui wi h he mul iplexe . (14) (15) 56 𝐹𝑖𝑙𝑡𝑒𝑟𝐹𝑟𝑒𝑞𝐶𝑀 = 1 2𝜋𝑅𝐶𝑐 whe e, 𝐶𝑑>10𝐶𝑐. The schema ic o he ci cui is shown in igu e 16. The inal e sion o he eading ci cui can be seen in igu e 17. Figu e 17 – Schema ic ep esen a ion o he inal ci cui . Figu e 16 – Inpu il e ha p e en s RF ampli ica ion e o s and low pass- il e . (16) 63 As can be seen in igu e 28 his g aphical in e ace is e y simple and use - iendly. Ini ially, i opens he i s window o he de ini ion o a iables and RTD selec ion. I was also added some ins uc ions o help he use unde s and how o use i . On he i s window, he use connec s he boa d, chooses he senso ha wan s o ead, inse s and sa es he a iables ha will be used in he con e sion and s a s he eading. A e his selec ion, a second window is open o show he empe a u e da a ha was con e ed and whe e he use can s op he eading and disconnec he boa d. Figu e 27 - S a e machine o he sys em. Figu e 28 - Implemen ed g aphical in e ace o choose and p esen he RTD esponse. 64 The sys em s a s, as ese mode, in a s a e called “Disconnec ed” whe e he mic ocon olle is no eading alues and is no connec ed o he compu e The connec ion be ween he compu e and he mic ocon olle is ob ained by p essing he bu on “connec ”. I he boa d is no connec ed co ec ly, an e o message will appea and he sys em will s ay in a s a e o disconnec ion. On o he hand, i he connec ion is success ul, a message will appea o in o m ha he connec ion was co ec ly pe o med ( igu e 29). Once he connec ion is made wi h success, he nex s ep o he use is o choose he senso o ead and se he a iables ha will ake pa in he con e sion o he ol age alues o empe a u e. The use canno s a he eading wi hou inse ing alues ha a e nume ic and di e en om ze o, as i shows an e o message ( igu e 30 a). When he a iables a e all se , he use p esses he sa e bu on o sa e he alues o he a iables and o send he numbe o he RTD ha is going o be ead by he mic ocon olle . As he a iables a e success ully sa ed, a con i ma ion message will appea on he sc een ( igu e 30 b). To ese he a iables inse ed, he use mus p ess he ese bu on o e ase he p e ious alues and inse he new ones. Once he connec ion is made wi h success, he nex s ep o he use is o choose he senso o ead and se he a iables ha will ake pa in he con e sion o he ol age alues o empe a u e. The use canno s a he eading wi hou inse ing alues ha a e nume ic and di e en om ze o, as i shows an e o message ( igu e 30 a). When he a iables a e all se , he use p ess he sa e bu on o sa e Figu e 30 - Wa nings ela ed o a iables inse ion. Figu e 29 - Connec ion wa nings. a b 65 he alues o he a iables and o send he numbe o he RTD ha is going o be ead by he mic ocon olle . P essing s a , he mic ocon olle ini ia es he eading and sends he alues ia UART. The p og am ecei es he ol age alues which a e con e ed o empe a u e and hen a e displayed in a g aphical o m and on a digi al LCD. These con e ed alues a e also sa ed in a ex ile. Each senso has a ex ile o ol age and empe a u e. The use can dele e all he iles by p essing he “dele e iles” bu on, o he wise, e e y eading is s o ed in he same ex ile o he p e ious eading. The use s ill has he possibili y o see he empe a u e o he ou senso s using he “All senso s” mode. This mode will change he senso ha is being ead e e y 30 sec and sa e all he con e ed da a in di e en iles, one o each senso . Finally, he use has wo op ions: 1) p ess s op and he sys em only s ops eading and con e ing alues o 2) disconnec all he sys em going back o he ese s a e o he sys em, by p essing he disconnec bu on. 4.3.2 Liquid C ys al Display (LCD) As p e iously men ioned, he sys em has a ious s a es and allow selec he RTD ha is being ead. This in o ma ion will be displayed on an LCD. Fo his pu pose, he LCD DFR0063 16x02 wi h communica ion ia I2C p o ocol was used. The I2C p o ocol uses 2 wi es, SCL (Se ial Clock) and SDA (Se ial Da a) o communica e be ween he mas e , in his case, he mic ocon olle and he sla e, in his case, he LCD. Fi s ly, he mas e sends a s a condi ion ollowed by eigh bi s whe e se en a e he sla e add ess and he eigh h bi is he mode ha he communica ion is going o ope a e, 0 is o w i e and 1 o ead om he sla e. I wai s o he acknowledge signal om he sla e ha he connec ion was made success ully. Then he mas e sends he da a, one by e a a ime and o each by e mus wai o an acknowledged signal ha he by e was well- ecei ed be o e sending he nex by e o da a. This ope a ion 66 con inues in his o m un il he mas e sends a s op condi ion o in e up he communica ion. In igu e 31, i can be seen he schema ic o he communica ion. In igu e 32, i can be seen all he s a es ha he LCD can show. Figu e 31 - I2C communica ion p o ocol schema ic. Figu e 32 - LCD showing he ou s a es o he sys em. 67 5. TESTS AND RESULTS In his chap e , he expe imen al es s and consequen esul s o he p ojec a e p esen ed and discussed. The de elopmen o he elec onic ci cui s in ol ed se e al es s eps namely in 1) b eadboa d, 2) p o oboa d and 3) PCB es ing. Imp o emen s in he elec onics ci cui s a e also p esen ed in his chap e . 5.1.1 B eadboa d es ing Fi s ly, he ci cui was implemen ed in he b eadboa d (Figu e 33), wi hou il e and wi h a gain o a ound 11 (as men ioned in sec ion 4.1.4). This es ing was made o alida e he senso esponse using a ho pla e wi h a esolu ion o 0.1°C o change he empe a u e. I was clea , as i can be seen in igu e 34, ha he senso was esponding as expec ed, i s ou pu ol age inc eased wi h a sensi i i y o 0.73 mV/°C. Howe e , he sys em s ill had some pe u ba ion de i a ed om noise and p esen a poo esolu ion. Figu e 34 - Mic osenso esponse wi h implemen ed elec onics o empe a u e a ia ion om 35 o 45 °C wi hou il e ing. Figu e 33 - Ci cui implemen ed in he b eadboa d. 68 A e his, a il e wi h a cu ing equency o 106 kHz was implemen ed o pe mi he eading o low- equency signals and a enua es equencies highe han he cu ing equency. As can be seen in igu e 35, he noise was subs an ially educed, bu he ampli ica ion ci cui s ill has poo esolu ion since om 35°C o 45°C he ou pu ol age only changes a ound 10mV. Since i is in ended o ead he senso 's esponse wi h he STM, he ou pu ange mus change signi ican ly. Fu he mo e, he ou pu ol age s ill has a signi ican le el o noise, especially when is in ended o ead wi h a esolu ion o 0.1°C. 5.1.2 P o oboa d es ing The nex s ep was he implemen a ion o he ci cui in a P o oboa d which can be seen in igu e 36. This ci cui helped educe he noise om he b eadboa d acks and o alida e hese changes be o e implemen ing he PCB. This ci cui didn’ ha e ye he AD620 and he TMUX1104 because as a his s age he only objec i e was o alida e he esponse o he senso . This ci cui was used an ins umen a ion ampli ie INA126 o alida e he ampli ica ion sys em. Figu e 36 - Ci cui implemen ed in P o oboa d. Figu e 35 - Ou pu ol age o he senso wi h il e ing when he empe a u e is inc eased om 35 °C o 45 °C. 69 Also, he eading equency o he ADC was inc eased o 1kHz, which is equi alen o eading alues each 1 ms. In igu e 37, i can be seen he ou pu ol age o he ci cui in he p o oboa d when he empe a u e was inc eased om 35°C o 45°C. High le els o luc ua ions and noise a e e i ied in he ou pu ol age when his assemble was used. The P o oBoa d s ill p oduces noise om he connec ions be ween he componen s and, heo e ically, he INA126 is no he ideal componen compa ed o he AD620, as i p oduces mo e noise compa ing bo h cha ac e is ics [59,60]. Also, a digi al echnique o summing up he alues o 250 eadings and aking he a e age numbe o ha sum was implemen ed o emo e ou pu ol age picks o he senso esponse ( igu e 38). In his g aph, he e a e wo cycles: 1) he blue line ep esen s an inc ease o empe a u e om 35 o 45°C and 2) he o ange line dec ease empe a u e om 45 o 35°C. I can be no ed ha he ol age picks we e so ened as expec ed bu he e is s ill some noise and ins abili y ha comes om he P o oBoa d ci cui . A sensi i i y o 3.37mV/°C was ob ained wi h his app oach. Figu e 37 - Resul s wi h P o oBoa d ci cui . Hea ing om 35°C o 45°C. Figu e 38 - Resul s o a ull cycle o hea ing and cooling o he sys em om 35°C o 45°C in P o oboa d. 70 5.1.3 PCB es ing A e alida ion o he en i e ci cui wi h b eadboa d and p o oboa d, he PCB was designed and manu ac u ed wi h all he necessa y ci cui s. A e manu ac u e, he expe imen al se up shown in igu e 39 was used o cha ac e ize he esponse o he senso s using he PCB ci cui . Se e al hea ing and cooling cycles be ween 35 and 45°C we e applied o alida e he PCB ci cui . Figu e 40 shows one o hese es s wi h s abiliza ion a 40°C. The esponse was simila o p e ious es s, howe e , i is e i ied g ea e s abili y o he esponse wi h lowe luc ua ions and noise. A sensi i i y o 7.61 mV/°C was ob ained wi h his app oach. F om he cha ac e iza ion o he senso esponse, i was necessa y o implemen a s ep o con e sion o he ou pu ol age o he sys em in o empe a u e alues. Fo ha was used he equa ions men ioned in chap e 3, in subsec ion 3.1.1 This cha ac e iza ion was made using he compa ison me hod, compa ing he alues o he RTD agains he senso o he ho pla e. Figu e 39 - Expe imen al se up used o he empe a u e mic osenso cha ac e iza ion using he implemen ed PCB. Figu e 40 - Resul s om hea ing he sys em om 35°C o 45°C in PCB e sion. 71 In he de eloped so wa e i is possible o de ine speci ic a iables o each o he senso s o be measu ed, i.e., gain, R0 and α. The R0 and α alues we e ob ained om he measu emen o he elec ical esis ance alue o he RTD using a digi al mul ime e (Agilen 34410A 61/2). The measu ed esis ance alues a e shown in he g aph in igu e 41. F om hese alues, i was possible o de ine he calib a ion line and calcula e he R0 and R100, and consequen ly he α. So, he s aigh equa ion is he ollowing: 𝑅 = 3,5765 ∗ 𝑇 + 1511,3 and his ga e a alue o 𝑅0 o 1511,3 Ω and also a alue o 𝑅100 o 1869,09 Ω. Wi h hese wo alues, i was possible also calcula e he p ac ical alue o he empe a u e coe icien o esis i i y (α) using he ollowing equa ion: 𝛼 = 𝑅100 − 𝑅0 𝑅0 ∗ ∆𝑇 whe e R100 and R0 a e he alues o esis ance and ∆𝑇 is he di e ence o empe a u e, in his case, 100℃. The p ac ical alue o he coe icien o his senso was 0.002367432 Ω / Ω /℃. This di e ence in he alue can be explained by he ac ha hese senso s a e ab ica ed in hin ilm and he heo e ical alue o he pla inum was de ined o he bulk ma e ial. Figu e 41 – Cha ac e iza ion o he elec ical esis ance o he pla inum RTD. (18) ZZ) (19) 72 In o de o assess i he senso s a e able o measu e empe a u e changes o 0.1°C, was used a ho pla e o apply a hea ing cycle be ween 37.5 ℃ o 38.9 ℃ wi h an inc emen o 0.1°C. In igu e 42 a e showed he con e ed empe a u e alues o hese es s. Wi h hese esul s, i can be seen ha he senso is esponding p ope ly o he empe a u e inc emen s, despi e some di icul y o keep a cons an empe a u e in each s ep due o he ins abili y o he ho pla e (lack o isola ion and hea ing losses). Howe e , e en i is possible o see he a ia ions o 0.1°C, he ou pu ol age is s ill low o he esolu ion wan ed. 5.1.4 Imp o emen s o he ci cui Wi h he objec i e o inc easing he esolu ion o he sys em, was implemen ed a new app oach wi h wo s ages o ampli ica ion using he AD620. On his app oach, was gi en special a en ion o he e e ence ol age o bo h AD620 o op imize he ou pu o ange be ween 0 and 3.3V ( ol age alues measu ed by ADC) The i s s age o ampli ica ion has an RG = 1kΩ, which gi es a gain o a ound 50.4 and on he second s age, he RG is 7.8kΩ which gi es a gain o a ound 7.33. The combina ion o hese wo gains p o ides a o al gain o a ound 369.6. To assu e ha he ou pu anges be ween 0 and 3.3 o he desi ed empe a u e ange (35°C o 45°C) he e e ence o he i s s age was connec ed o -10V o allow he sys em o ha e a mo e ampli ica ion ange. Wi h he ou pu o he i s s age connec ed o he posi i e en ance o he second s age, he nega i e en ance can ma ch he ou pu alue o he empe a u e ha is wan ed o be he i s alue o he empe a u e eading. In his case, 23°C was chosen o be he ini ial alue. The ol age alue co esponding o his empe a u e is connec ed o he nega i e en ance and he ampli ie will sub ac bo h Figu e 42 - Resul s om he hea ing es wi h inc emen s o 0.1°C om 37.5°C o 38.9°C. 79 BIBLIOGRAPHY [1] J. E. Sosa-He nández e al., “O gans-on-a-Chip Module: A Re iew om he De elopmen and Applica ions Pe spec i e,” Mic omachines, ol. 9, no. 10, 2018, doi: 10.3390/mi9100536. [2] H. Mau iac, C. Panne ie , and G. V. Casquillas, “O gans on chip e iew,” El e low, Dec. 2020, Accessed: Dec. 19, 2021. [Online]. A ailable: h ps://www.el e low.com/mic o luidic- e iews/o gans-on-chip-3d- cell-cul u e/o gans-chip- e iew/. [3] Q. Wu e al., “O gan-on-a-chip: Recen b eak h oughs and u u e p ospec s,” Biomed. Eng. Online, ol. 19, no. 1, pp. 1–19, 2020, doi: 10.1186/s12938-020-0752-0. [4] B. A. McKenzie and W. H. G o e , “A mic o luidic he mome e : P ecise empe a u e measu emen s in mic oli e -and nanoli e -scale olumes,” PLoS One, ol. 12, no. 12, pp. 1–14, 2017, doi: 10.1371/jou nal.pone.0189430. [5] J. Kim, J. Kim, Y. Shin, and Y. Yoon, “A S udy on he Fab ica ion o an RTD (Resis ance Tempe a u e De ec o ) by Using P Thin Film,” Ko ean J. Chem. Eng., ol. 18, no. 1, pp. 61–66, 2001, doi: 10.1007/BF02707199. [6] J. Wu, “A Basic Guide o RTD Measu emen s,” Texas Ins umen s Inc., no. June, pp. 1–42, 2018. [7] O. Cann, “These a e he op 10 eme ging echnologies o 2016 | Wo ld Economic Fo um,” Wo ld Econ. Fo um, 2016, Accessed: No . 26, 2020. [Online]. A ailable: h ps://www.we o um.o g/agenda/2016/06/ op-10-eme ging- echnologies-2016/. [8] G. M. Whi esides, “The o igins and he u u e o mic o luidics,” Na u e, ol. 442, no. 7101, pp. 368–373, 2006, doi: 10.1038/na u e05058. [9] J. Chen, D. Chen, Y. Xie, T. Yuan, and X. Chen, “P og ess o mic o luidics o biology and medicine,” Nano- Mic o Le ., ol. 5, no. 1, pp. 66–80, 2013, doi: 10.3786/nml. 5i1.p66-80. [10] T. Chen e al., “A d ug-compa ible and empe a u e- con olled mic o luidic de ice o li e-cell imaging,” 2016. [11] M. Gonidec and J. Puigma í-Luis, “Con inuous- e sus segmen ed- low mic o luidic syn hesis in ma e ials science,” C ys als, ol. 9, no. 1, 2019, doi: 10.3390/c ys 9010012. [12] B. K. Gale e al., “Low-Cos MEMS Technologies,” Re . Modul. Ma e . Sci. Ma e . Eng., 2016, doi: 10.1016/B978-0-12-803581-8.00530-0. [13] W. Quispe-Tin aya, “乳鼠心肌提取 HHS Public Access,” Physiol. Beha ., ol. 176, no. 3, pp. 139–148, 2017, doi: 10.1039/c6lc00163g.The. [14] E. Roy, J.-C. Galas, and T. Ve es, “The moplas ic elas ome s o mic o luidics: owa ds a high- h oughpu ab ica ion me hod o mul ilaye ed mic o luidic de ices.,” Lab Chip, ol. 11, no. 18, pp. 3193–3196, Sep. 2011, doi: 10.1039/c1lc20251k. [15] X. Li, D. R. Balle ini, and W. Shen, “A pe spec i e on pape -based mic o luidics: Cu en s a us and u u e ends,” Biomic o luidics, ol. 6, no. 1, 2012, doi: 10.1063/1.3687398. [16] Y. Cong e al., “D ug oxici y e alua ion based on o gan-on-a-chip echnology: A e iew,” Mic omachines, ol. 11, no. 4, pp. 1–24, 2020, doi: 10.3390/MI11040381. [17] T. Kilic, F. Na aee, F. S adolini, P. Renaud, and S. Ca a a, “O gans-on-chip moni o ing: senso s and o he s a egies,” Mic ophysiological Sys ., ol. 1, no. 2, pp. 1–1, 2018, doi: 10.21037/mps.2018.01.01. [18] E. Ve poo e e al., “HOW MICROTECHNOLOGIES ENABLE ORGANS-ON-A-CHIP Pha maceu ical Analysis , 2 Pha macokine ics , Toxicology & Ta ge ing , G oningen Resea ch Ins i u e o Pha macy , Uni e si y o G oningen , G oningen , The Ne he lands Endo helial Biomedicine and Vascula D u,” pp. 224–227, 2015. [19] B. Zhang, A. Ko olj, B. F. L. Lai, and M. Radisic, “Ad ances in o gan-on-a-chip enginee ing,” Na . Re . Ma e ., ol. 3, no. 8, pp. 257–278, 2018, doi: 10.1038/s41578-018-0034-7. [20] B. J. Kane, M. J. Zinne , M. L. Ya mush, and M. Tone , “Li e -speci ic unc ional s udies in a mic o luidic a ay o p ima y mammalian hepa ocy es,” Anal. Chem., ol. 78, no. 13, pp. 4291–4298, 2006, doi: 10.1021/ac051856 . [21] D. Huh, B. Ma hews, A. Mammo o, M. Mon oya-Za ala, H. Hsin, and D. Ingbe , “Recons i u ing O gan- Le el Lung Func ions on a Chip,” ol. 328, 2010, [Online]. A ailable: h ps://science.sciencemag.o g/con en /328/5986/1662. [22] K.-J. Jang and K.-Y. Suh, “Minia u isa ion o chemis y, physics, biology, & bioenginee ing A mul i-laye mic o luidic de ice o e icien cul u e and analysis o enal ubula cells †,” ol. 10, pp. 1–132, 2010, doi: 10.1039/b907515a. [23] A. G osbe g, A. P. Nesmi h, J. A. Goss, M. D. B igham, M. L. McCain, and K. K. Pa ke , “Muscle on a chip: 80 In i o con ac ili y assays o smoo h and s ia ed muscle,” J. Pha macol. Toxicol. Me hods, ol. 65, no. 3, pp. 126–135, May 2012, doi: 10.1016/j. ascn.2012.04.001. [24] Y. Imu a, Y. Asano, K. Sa o, and E. Yoshimu a, “A mic o luidic sys em o e alua e in es inal abso p ion.,” Anal. Sci. In . J. Japan Soc. Anal. Chem., ol. 25, no. 12, pp. 1403–1407, Dec. 2009, doi: 10.2116/analsci.25.1403. [25] P. M. an Midwoud, M. T. Me ema, E. Ve poo e, and G. M. M. G oo huis, “A mic o luidic app oach o in i o assessmen o in e o gan in e ac ions in d ug me abolism using in es inal and li e slices,” Lab Chip, ol. 10, no. 20, pp. 2778–2786, 2010, doi: 10.1039/C0LC00043D. [26] N. F anzen, W. H. an Ha en, V. P. Re èl, P. Loskill, J. an den Eijnden- an Raaij, and M. IJze man, “Impac o o gan-on-a-chip echnology on pha maceu ical R&D cos s,” D ug Disco . Today, ol. 24, no. 9, pp. 1720–1724, 2019, doi: 10.1016/j.d udis.2019.06.003. [27] W. H. an Ha en, A. Wind, P. de Paoli, M. Sagha chian, and S. Obe s , “Ac ual cos s o cance d ugs in 15 Eu opean coun ies,” Lance Oncol., ol. 17, no. 1, pp. 18–20, Jan. 2016, doi: 10.1016/S1470- 2045(15)00486-6. [28] E. Alza e, “Medición de empe a u a : senso es e moeléc icos,” no. June, 2016. [29] I. Te asaki, “The mal Conduc i i y and The moelec ic Powe o Semiconduc o s,” Re . Modul. Ma e . Sci. Ma e . Eng., 2016, doi: 10.1016/B978-0-12-803581-8.00771-2. [30] A. Alashka and A. H. Alami, “O e iew o The moelec ic Ma e ials,” in Encyclopedia o Sma Ma e ials, Else ie , 2022, pp. 319–325. [31] L. Mo ei a, “Medição de Tempe a u a Usando-se Te mopa ,” Ce âmica Ind., ol. 7, no. 5, pp. 51–53, 2002, [Online]. A ailable: h p://www.ce amicaindus ial.o g.b /pd / 07n05/ 7n5_6.pd . [32] “O e iew o Tempe a u e Senso s - NI.” h ps://www.ni.com/p -p /inno a ions/whi e- pape s/06/o e iew-o - empe a u e-senso s.h ml (accessed Oc . 20, 2020). [33] C. F. Tsai, L. T. Li, C. H. Li, and M. S. Young, “Implemen a ion o he mis o linea iza ion using LabVIEW,” IIH-MSP 2009 - 2009 5 h In . Con . In ell. In . Hiding Mul imed. Signal P ocess., ol. 1, pp. 530–533, 2009, doi: 10.1109/IIH-MSP.2009.98. [34] S. Mokha ab, W. A. Poe, and J. Y. Mak, “Gas P ocessing Plan Au oma ion,” Handb. Na . Gas T ansm. P ocess., pp. 615–642, 2019, doi: 10.1016/B978-0-12-815817-3.00020-4. [35] J. Y. Hu, Y. I. Li, and K. W. Yeung, “In i o physiological measu emen s,” Clo h. Biosenso y Eng., pp. 285– 300, 2006, doi: 10.1533/9781845691462.285. [36] V. V. Gu eye and A. A. L’ o , “High accu acy semiau oma ic calib a ion o indus ial RTDs,” Con . Rec. - IEEE Ins um. Meas. Technol. Con ., pp. 1–5, 2007, doi: 10.1109/im c.2007.379291. [37] V. K. Na ajan and K. T. Ch is ensen, “Tempe a u e Measu emen , Me hods,” in Encyclopedia o Mic o luidics and Nano luidics, D. Li, Ed. Bos on, MA: Sp inge US, 2008, pp. 1994–2005. [38] S. A. Binsilm, M. A. Rad, and M. R. Ahmad, “Modeling o in eg a ed nanoneedle-mic o luidic sys em o single cell empe a u e measu emen ,” Appl. Sci., ol. 6, no. 12, 2016, doi: 10.3390/app6120339. [39] J.-M. Yang, H. Yang, and L. Lin, “Quan um Do Nano The mome e s Re eal He e ogeneous Local The mogenesis in Li ing Cells,” ACS Nano, ol. 5, no. 6, pp. 5067–5071, 2011, doi: 10.1021/nn201142 . [40] J. Lee, A. O. Go o o , and N. A. Ko o , “Nanopa icle assemblies wi h molecula sp ings: A nanoscale he mome e ,” Angew. Chemie - In . Ed., ol. 44, no. 45, pp. 7439–7442, 2005, doi: 10.1002/anie.200501264. [41] C. Espenel, M. C. Giocondi, B. Sean ie , P. Dosse , P. E. Milhie , and C. Le G imellec, “Tempe a u e- dependen imaging o li ing cells by AFM,” Ul amic oscopy, ol. 108, no. 10, pp. 1174–1180, Sep. 2008, doi: 10.1016/j.ul amic.2008.04.090. [42] C. Li, P. M. Wu, J. Han, and C. H. Ahn, “A lexible polyme ube lab-chip in eg a ed wi h mic osenso s o sma mic oca he e ,” Biomed. Mic ode ices, ol. 10, no. 5, pp. 671–679, 2008, doi: 10.1007/s10544- 008-9178-3. [43] N. Inoma a, M. Toda, M. Sa o, A. Ishijima, and T. Ono, “Pico calo ime e o de ec ion o hea p oduced in an indi idual b own a cell,” Appl. Phys. Le ., ol. 100, no. 15, 2012, doi: 10.1063/1.3701720. [44] P. Schmi , K. Wed ich, L. Mülle , H. Mehne , and M. Ho mann, “Design, ab ica ion and cha ac e isa ion o a mic o luidic ime-Tempe a u e indica o ,” J. Phys. Con . Se ., ol. 922, no. 1, 2017, doi: 10.1088/1742-6596/922/1/012004. [45] J. Oile and H. Yu, “The moelec ic Cool- ilm Flow Senso ,” 2012. [46] C. Li, J. Sun, Q. Wang, W. Zhang, and N. Gu, “Wi eless The mome y o Real-Time Tempe a u e Reco ding on Thousand-Cell Le el,” IEEE T ans. Biomed. Eng., ol. 66, no. 1, pp. 23–29, 2019, doi: 10.1109/TBME.2018.2836949. 81 [47] J. Qiao, X. Mu, and L. Qi, “Cons uc ion o luo escen polyme ic nano- he mome e s o in acellula empe a u e imaging: A e iew,” Biosens. Bioelec on., ol. 85, pp. 403–413, 2016, doi: h ps://doi.o g/10.1016/j.bios.2016.04.070. [48] Y. G. Shellman e al., “Fas esponse empe a u e measu emen and highly ep oducible hea ing me hods o 96-well pla es,” Bio echniques, ol. 36, no. 6, pp. 968–976, 2004, doi: 10.2144/04366ST01. [49] D. J. King, “Resis ance Elemen s and RTD ’s,” [Online]. A ailable: h p://www.omega.com/ empe a u e/Z/pd /z054-056.pd . [50] A. Komola e, R. To ah, H. Nunes-Ma os, M. Tudo , and S. Beeby, “In eg a ion o empe a u e senso s in ab ics,” FLEPS 2019 - IEEE In . Con . Flex. P in able Senso s Sys . P oc., 2019, doi: 10.1109/FLEPS.2019.8792294. [51] E. J. P. San os and I. B. Vasconcelos, “RTD-based sma empe a u e senso : p ocess de elopmen and ci cui design,” “2008 26 h In . Con . Mic oelec on. P oceedings, MIEL 2008,” no. Miel, pp. 333–336, 2008, doi: 10.1109/ICMEL.2008.4559289. [52] M. S. Abo Ghazala, H. A. O hman, L. M. Sha a El-Deen, M. A. Nawwa , and A. E. H. B. Kashyou , “Fab ica ion o nanoc ys alline silicon hin ilms u ilized o op oelec onic de ices p epa ed by he mal acuum e apo a ion,” ACS Omega, ol. 5, no. 42, pp. 27633–27644, 2020, doi: 10.1021/acsomega.0c04206. [53] S. Li e al., “Tunable op ical p ope ies o amo phous-like ga2o3 hin ilms deposi ed by elec on-beam e apo a ion wi h a ying oxygen pa ial p essu es,” Nanoma e ials, ol. 10, no. 9, pp. 1–10, 2020, doi: 10.3390/nano10091760. [54] M. El-Dai i and R. J. House, “Op ic ne e hypoplasia,” Handbook o Pedia ic Re inal OCT and he Eye- B ain Connec ion. pp. 285–287, 2019, doi: 10.1016/B978-0-323-60984-5.00062-7. [55] H. Tanaka, “Epi axial G ow h o Oxide Films and Nanos uc u es,” in Handbook o C ys al G ow h: Thin Films and Epi axy: Second Edi ion, ol. 3, Else ie Inc., 2015, pp. 555–604. [56] L. Wei, X. Shusheng, and Z. Xiaojun, “Lead-Wi e-Resis ance Compensa ion Technique,” Senso s, ol. 20,2742, 2020. [57] T. Junsing, “In e ace Ci cui o Th ee-Wi e Resis ance Tempe a u e De ec o wi h Lead Wi e Resis ance Compensa ion,” no. 2, pp. 1–4, 2020, doi: 10.1109/ i2c48728.2019.8999909. [58] K. E kan, B. Akkoyunlu, E. Balkan, and M. Tayanç, “A po able bo ehole empe a u e logging sys em using he ou -wi e esis ance me hod,” J. Geophys. Eng., ol. 14, no. 6, pp. 1413–1419, 2017, doi: 10.1088/1742-2140/aa7 e. [59] P. J. Sousa, V. C. Pin o, V. H. Magalhães, R. O. Rod igues, P. C. Sousa, and G. Minas, “De elopmen o highly sensi i e empe a u e mic osenso s o localized measu emen s,” Appl. Sci., ol. 11, no. 9, 2021, doi: 10.3390/app11093864. [60] P ecise, “The mome e Calib a ion by he Fixed Poin Me hod - PTS.” h p://p ecisecalib a ions.com/ he mome e -calib a ion- ixed-poin -me hod/ (accessed Dec. 30, 2020). [61] O. Soa, “L79 da ashee ,” STMic oelec onics, no. Janua y, 2018, [Online]. A ailable: h ps://www.s .com/ esou ce/en/da ashee /l79.pd . [62] F. P. Dpak, “Posi i e ol age egula o ICs,” no. Feb ua y, pp. 1–59, 2016. [63] S. Lm, “LM134 / LM234 / LM334 3-Te minal Adjus able Cu en Sou ces,” no. Ma ch 2000, 2013. [64] D. In o ma ion, A. Example, and B. Diag am, “TMUX1108 5V / ± 2 . 5V , Low-Leakage-Cu en , 8 : 1 P ecision Mul iplexe ,” 2018. [65] A. De ices, “AD620 Low Cos , Low Powe Ins umen a ion Ampli ie ,” Online] www. da ashee ca alog. com.[Online], pp. 1–20, 1999, [Online]. A ailable: h p://schola .google.com/schola ?hl=en&b nG=Sea ch&q=in i le:Low+Cos +Low+Powe +Ins umen a ion+Ampli ie +AD620#0%5Cnh p://schola .google.com/schola ?hl=en&b nG=Sea ch&q=in i le:Low+C os ,+Low+Powe +Ins umen a ion+Ampli ie +AD620#0. [66] INAx126, “Mic oPowe Ins umen a ion Ampli ie Single and Dual Ve sions,” Texas Ins umen s, p. 34, 2015, [Online]. A ailable: h ps:// inyu l.com/y228 xyw.