1
Scien i ic RepoR S | (2020) 10:6925 | h ps://doi.o g/10.1038/s41598-020-63620-5
www.na u e.com/scien i ic epo s
Rapid cha ac e iza ion o
Biomolecules’ he mal S abili y
in a Segmen ed low- h ough
Op o luidic Mic osys em
Zdenka ohle o a1,2, Hanliang Zhu3, Ja omi Hubalek1,2, Sheng ni4, Le en Yobas4,
pa el podes a1, Alexand o ahal2 & pa el neuzil1,2,3 ✉
Op o luidic de ices combining op ics and mic o luidics ha e ecen ly a ac ed a en ion o
biomolecula analysis due o hei high de ec ion sensi i i y. He e, we show a silicon chip wi h
ubula mic ochannels bu ied inside he subs a e ea u ing empe a u e g adien (∇T) along he
mic ochannel. We se up an op ical luo escence sys em consis ing o a powe -modula ed lase ligh
sou ce o 470 nm coupled o he mic ochannel se ing as a ligh guide ia op ical ibe . Fluo escence was
de ec ed on he o he side o he mic ochannel using a pho omul iplie ube connec ed o an op ical
ibe ia a luo escein iso hiocyana e il e . The PMT ou pu was connec ed o a lock-in ampli ie o
signal p ocessing. We pe o med a mel ing cu e analysis o a sho dsDNA – SYBR G een I complex
wi h a known mel ing empe a u e (TM) in a low- h ough con igu a ion wi hou g adien o e i y he
unc ionali y o he p oposed de ec ion sys em. We hen used he segmen ed low con igu a ion and
measu ed he luo escence ampli ude o a d ople exposed o ∇T o ≈ 2.31 °C mm−1, de e mining
he hea ans e ime as ≈ 554 ms. The p oposed pla o m can be used as a as and cos -e ec i e
sys em o pe o ming ei he MCA o dsDNAs o o measu ing p o ein un olding o d ug-sc eening
applica ions.
Tempe a u e can signi ican ly a ec he biological sys ems o li ing o ganisms in e ms o cellula mo phology,
me abolism, g ow h, and cell dea h1,2. On a molecula le el, empe a u e in luences he s uc u e and unc ion
o biomolecules such as p o eins and nucleic acid3,4. The dena u a ion o biomolecules unde ela i ely high em-
pe a u es is u ilized in eal- ime polyme ase eac ions (qPCR), subsequen mel ing cu e analyses (MCA) o
double-s anded deoxy ibonucleic acid (dsDNA), and di e en ial scanning luo ime y (DSF) o p o eins.
The MCA analysis o dsDNA s ained wi h an in e cala o dye such as SYBR G een I is conduc ed by g adu-
ally amping he empe a u e while moni o ing he in ensi y o he emi ed luo escence (F). The dsDNA mel -
ing empe a u e (TM) is de ined as he empe a u e a which 50% o p esen ed dsDNA molecules sepa a e in o
single-s anded o ms. This alue depends on he numbe o base pai s and hei composi ion. Typically, while
he MCA is pe o med using he qPCR sys ems, hey can be also used o de e mine he p o ein-un olding cu e.
Howe e , qPCR sys ems cu en ly a ailable a e bulky, cos ly, and slow.
Ad ances in mic o ab ica ion echnology ha e enabled he minia u iza ion o bioanaly ical ools wi h p e-
cisely con olled empe a u es, allowing o quick in es iga ions in o he he mal e ec s on biological molecules
using samples wi h a small olume. Hea ing me hods and he he mal conduc i i y o he de ice’s ma e ials a e
essen ial o c ea e he mally s able sys ems5 wi h ei he homogeneous dis ibu ion o empe a u e o empe a u e
g adien (∇T).
Hea conduc ion-based mic o luidic echniques, especially in a d ople con igu a ion, ha e a ac ed a en-
ion due o hei applicabili y in DNA ampli ica ion, p o ein analysis, single-cell assays, and chemical syn hesis.
1Cen al Eu opean Ins i u e o Technology, B no Uni e si y o Technology, Pu kyno a 123, 612 00, B no, Czech
Republic. 2Depa men o Mic oelec onics, Facul y o Elec ical Enginee ing and Communica ion, B no Uni e si y
o Technology, Technicka 3058/10, 61600, B no, Czech Republic. 3Minis y o Educa ion Key Labo a o y o Mic o and
Nano Sys ems o Ae ospace, School o Mechanical Enginee ing, No hwes e n Poly echnical Uni e si y, 127 Wes
Youyi Road, Xi’an, Shaanxi, 710072, P.R. China. 4Hong Kong, Uni e si y o Science and Technology, Clea Wa e Bay,
Hong Kong, P.R. China. ✉e-mail: [email p o ec ed]
open
2
Scien i ic RepoR S | (2020) 10:6925 | h ps://doi.o g/10.1038/s41598-020-63620-5
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
The cha ac e is ics o he he mal beha io o wa e in oil d ople s wi hin mic o luidic channels we e s udied
pe he gene a ion and anspo o liquid d ople s wi hin local hea ing a he b eakup loca ion o T-junc ions
and low- ocusing sys ems6,7 ei he a he downs eam channel8,9 o wi h he hea e placed a he end o he
anspo channel10. Howe e , he empe a u e dependency o d ople physical p ope ies ep esen ed by den-
si y, iscosi y, and in e acial ension be ween he con inuous oil phase and wa e -based d ople may complica e
high-pe o mance d ople mic o luidics. Fu he , he ma hema ical simula ion o d ople beha io unde hea ing
is complica ed due o he necessi y o 3D models o a d ople -based sys em11. In addi ion o he in es iga ion o
he empe a u e dependence o a d ople ’s physical p ope ies wi h localized hea ing on he mic o luidic chip,
he c ea ion o on-chip ∇T has been challenging as one can s udy mic o ubule polyme iza ion12 along he ∇T
o MCA o single nucleo ide polymo phisms13, bo h pe o med in a low- h ough con igu a ion. Mo eo e , he
MCA in mic o luidics has been conduc ed using ei he solid phase mul iple analysis, which equi es DNA immo-
biliza ion13,14, o he immobiliza ion o a ee liquid phase es ic ed o a single analysis15. Howe e , mic o-/nano-
scale dimensions o channels and he su ounding en i onmen and low- h ough con igu a ion may complica e
he de e mina ion o he empe a u e dis ibu ion in he mally d i en mic ode ices16. Thus, hea ans e in such
a sys em should be cha ac e ized in hea luxes by he mal con ec ion, he mal adia ion, o he hea ans e ime
be ween he sidewall o a chip and liquid17,18.
Op ical sys ems ha e been used as a de ec o o analyzing bio/chemical samples as well as hei minia u i-
za ion u ilized in lab-on-a-chip (LOC) applica ions19 such as po able qPCR o DNA20 o RNA diagnos ics21,22.
Recen ly, he combina ion o op ics and mic o luidics has a ac ed g ea a en ion due o he combina ion o
highly sensi i e bio-de ec ion wi h LOC echnology23,24. By ully in eg a ing op ical unc ions on a chip ins ead
o using bulky ex e nal op ics, hese op o luidic sys ems lowe he cos and downsize he sys em, making i p om-
ising o poin -o -ca e diagnosis25. The de ec ion o luo escence emains aluable in many bio-applica ions.
The e o e, he de elopmen and cha ac e iza ion o a mul i unc ional op o luidic lab-on-a-chip was in oduced
o sample analysis by luo escence and Raman spec oscopy26 o by he measu emen o he abso bance and
luo escence o d ople s in segmen ed low27. Fu he , he luo escence de ec ion and quan i ica ion o he Ebola
i us using hyb id op o luidic in eg a ion ha e been epo ed28, as ha e pla o ms o eal- ime isualiza ion o
i uses in complex media29 o cell phone-based imaging cy ome e s30.
He e, we demons a e he MCA o dsDNA wi h he sensi i e de ec ion o luo escence inside a mic ochan-
nel as pa o an op ical mic osys em exposed o ∇T alue. We used a modula ed lase as he ligh sou ce and a
pho omul iplie ube (PMT) as he luo escence de ec o wi h i s ou pu signal p ocessed by a lock-in ampli ie
o supp ess en i onmen al noise. The p oposed concep o a he mally egula ed op o luidic pla o m was i s
cha ac e ized in low- h ough con igu a ion o dsDNA bulk. Mo e impo an ly, we p esen ed he MCA analysis
in segmen ed low con igu a ion as a ool o apid de e mina ion o a biomolecules’ s abili y.
Ma e ials and Me hods
chip ab ica ion. We designed he chip layou wi h a size o (6 × 27) mm2 using he Nanoli hog aphy ool-
box so wa e31 wi h he aim o ha ing all luid/op ics inpu s/ou pu s a he chip sidewalls o p o ide a obus
solu ion. The layou consis ed o ≈30 µm-wide lines subsequen ly o ming bu ied mic ochannels by a p ocess
simila o ea lie ones ha used wo pa ylene-C deposi ions32. The i s pa ylene-C was employed as a mask wi h
con o mal coa ing on he sidewalls and a second pa ylene-C laye o seal he c ea ed channels33. We designed ou
de ice o inse bo h capilla ies and he op ical ibe s om he sides (Fig.1A), allowing a mo e obus con igu a-
ion han he p e ious one, which had i s capilla ies inse ed e ically.
We s a ed he ab ica ion p ocess using Si wa e s wi h a diame e o ≈ 100 mm and an unusual hickness: ≈
1 mm, which is hicke han he o dina y wa e hickness o ≈ 450 µm o 550 µm since he a ge ed mic ochannel
diame e was he same: be ween ≈ 450 µm and ≈ 550 µm.
Fi s , we coa ed he wa e s wi h a ≈ 10 µm- hick posi i e pho o esis (PR), ollowed by a p ebake a ≈ 110 °C
o ≈ 165 s. The wa e s we e exposed o ul a iole ligh wi h a dose o ≈ 1600 mJ·cm−2 o li hog aphy and de el-
oped in a KOH-based de elope o ≈ 300 s, pa e ning he shape o he enches a a designed wid h o 30 µm.
The Si was e ched wi h he Bosch p ocess34 o a a ge dep h o ≈ 500 µm, hen, we s ipped PR in N-me hyl-
2-py olidinone solu ion a ≈ 80 °C o ≈ 600 s (Fig.1B).
The wa e was hen coa ed wi h a ≈ 1.5 µm pa ylene-C laye (Fig.1C), and his laye was pho o-blas ed a he
ench bo oms using a em osecond lase wo king a a p incipal wa eleng h o ≈ 515 nm using pulses wi h a
du a ion o ≈ 300 s and a maximum ampli ude o pulse ene gy o ≈ 200 μJ (Fig.1D). The wa e was exposed o
XeF2 apo o iso opically e ch Si h ough he opening in he ench bo om, o ming a bu ied cylind ical mic o-
channel wi h a diame e o ≈ 500 µm (Fig.1E). The pa ylene-C was emo ed using O2 plasma wi h a se powe o
300 W o a du a ion o 1 h (Fig.1F). The wa e was diced in o indi idual chips, and he second pa ylene-C laye
was deposi ed wi h a hickness o ≈ 30 µm o co e he mic ochannel wi h hyd ophobic ma e ials and o seal he
≈ 30 µm-wide ench (Fig.1G). Bo h op ical ibe s and mic ocapilla ies we e inse ed in o he chip and sealed
wi h epoxy esin. De ails o one po a e shown in Fig.1H.
Op o luidic. The Si chip had a bu ied mic ochannel sys em consis ing o a h ough channel and h ee inle s
connec ed ia a c oss junc ion and a single ou le . The h ough channel also se ed as a ligh guide wi h an op o-
luidic pa h wi h a leng h o ≈ 24 mm (Fig.2B). The diame e o he bu ied mic ochannel was chosen o be ≈
500 µm (Fig.2A), su icien ly la ge o bo h he op ical ibe and he capilla y o be inse ed inside he chip o o m
a s able, obus , and au onomous sys em. The chip was placed on wo b ass blocks wi h a dis ance o ≈ 13 mm
(Fig.2C). Each block had i s own hea e and senso connec ed o a p opo ional in eg a i e de i a i e empe -
a u e con olle . The hea e s’ empe a u e was se o ≈ 50 °C and ≈ 80 °C, espec i ely, o ming a ∇T alue o ≈
2.31 °C·mm−1. We inse ed wo mul imode op ical ibe s wi h a co e and cladding diame e o (200 ± 4) µm and
(220 ± 2) µm, espec i ely, bo h (mean ± s anda d de ia ion) in o he h ough mic ochannel o he chip. Each
3
Scien i ic RepoR S | (2020) 10:6925 | h ps://doi.o g/10.1038/s41598-020-63620-5
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
o he op ical ibe s was placed on an opposi e side o he chip and sealed wi h epoxy esin. One ibe was con-
nec ed o lase -p oducing ligh wi h a p incipal wa eleng h o 471 nm and a nominal powe o 1 W. I s powe was
elec ically modula ed sinusoidal AC ol age om he lock-in ampli ie ’s in e nal powe supply wi h a equency
and ampli ude se o 1.2345 kHz and 1.958 V, espec i ely. The lase powe en e ing he op o luidic chip was a en-
ua ed om i s o iginal alue o 1 W by a se o neu al densi y il e s o ≈ 25 mW. The ligh lea ing he chip was
coupled in o a second op ical ibe connec ed ia a bandpass il e wi h a cen e wa eleng h and bandwid h o ≈
525 nm and ≈ 50 nm, espec i ely, o block he second ha monic lase ligh wi h a wa eleng h o ≈ 471 nm and
ha monic o ≈ 942 nm. We used a PMT as a ligh de ec o , ha ing i s gain se o ≈ 5 × 104 by se ing he con ol
ol age o 0.67 V. The PMT ou pu was connec ed o he lock-in ampli ie inpu wi h sensi i i y and ime cons an
se o 50 mV pe ull ange and 300 ms, espec i ely.
The high powe o he lase in e ac ing wi h he luid inside he mic ochannel could a ec he measu emen
as he powe co esponds o he powe densi y o ≈ 127 mW·mm−2. Ne e heless, he pho o he mal e ec should
Figu e 1. (A) Simpli ied iew o whole chip wi h op ical ibe s inse ed on he sho edges o he chip acing
each o he and wo capilla ies inse ed pa allel in o he long edge o he chip. (B) The Si subs a e was pa e ned
and e ched using he Bosch p ocess o o m enches a a wid h o 30 µm. (C) The pa ylene-C wi h hickness o
≈ 1.5 µm was deposi ed, and (D) he ench loo was la e pho o-blas ed by em osecond o expose he silicon.
(E) The wa e was subjec ed o XeF2 apo s, emo ing silicon om he ench loo and hus o ming a ubula
channel wi h a diame e o ≈500 µm. (F) Pa ylene was s ipped o using O2 plasma. (G) The wa e s we e diced
in o indi idual chips, and a second laye o pa ylene-C was deposi ed wi h a hickness o ≈ 30 µm sealing he
enches and coa ing sidewalls o he channels. (H) As he las s ep, he capilla y and op ical ibe s we e inse ed
in o he side holes and sealed wi h an epoxy esin.
Figu e 2. (A) The designed mic o luidic chip had a size o ≈ (6 × 27) mm2. The layou consis ed o wo le els:
he ed and b own colo s ep esen he channels and he h ough-holes, espec i ely. (B) Pho og aph o a
ab ica ed chip made o silicon (bo om) capped. (C) Close-up pho og aph o he chip wi h hea e s, senso s,
a ached op ical ibe s and capilla ies, and (D) block diag am o a comple e es ing se up wi h simpli ied
op o luidic chip layou , whe e VL is he ou pu signal om he lock-in ampli ie .
4
Scien i ic RepoR S | (2020) 10:6925 | h ps://doi.o g/10.1038/s41598-020-63620-5
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
no be e y se e e. Fi s , he oil/wa e in e ac s wi h ligh a a a he low nominal wa eleng h o 471 nm, bu
he e is, o cou se, an in e ac ion wi h luo escence-p oducing dyes, such as SYBR G een I in p esence o dsDNA,
which wa ms he wa e -based sample. The pho o he mal e ec was de e mined ea lie using 500 imes mo e
powe ul lase pe uni o a ea, which only wa med he wa e by ≈ 6 K35. Finally, we pe o med a ela i e meas-
u emen , no absolu e, and hus hea ing o he sample by pho o he mal e ec is compensa ed.
The luid was connec ed o he chip ia silica capilla ies wi h in e nal and ex e nal diame e s o ≈ 100 µm
and 360 µm, espec i ely, o inpu s and ou pu , which we e sealed wi h epoxy esin (Fig.2D). We used a
p essu e-con olled sys em o con ol he luid low a e (ν). The chip was e en designed o be able o gene a e
segmen ed low using a c oss-junc ion; we o med i ex e nally using a double T-junc ion as be o e36, since he
ex e nally gene a ed segmen ed low was mo e s able han he one o med in e nally.
chemicals. Syn he ically p epa ed dsDNA has a leng h o 17 bps wi h a sequence o 5’-TCT GCT GTC ACA
ACT AA-3’. The TM alue o he dsDNA is ≈ 60.6 °C as e i ied by a comme cial qPCR sys em. We p epa ed he
es solu ion by mixing his dsDNA a a concen a ion o ≈ 20 nM wi h SYBR-G een I dilu ed 10,000 imes in
a T is-EDTA bu e (1×). We also used his bu e wi h no dsDNA as a e e ence. An oil phase o hexadecane
supplemen ed wi h 2% o SPAN-80 su ac an was used as an immiscible con inual phase o segmen ed low
analysis.
Resul s and Discussion
Ma hema ical calcula ion o hea dis ibu ion. De ices ope a ing a ele a ed ai empe a u es wi h
he sample lowing h ough a e subjec o hea losses by con ec ion, adia ion, and hea lux due o sample low,
possibly a ec ing desi ed empe a u e dis ibu ion as well as i s ∇T37,38. We will only b ie ly desc ibe he physical
analysis, i s nume ical modeling, and expe imen al e i ica ion by in a ed (IR) imaging. De ails o hose h ee
me hods we e ecen ly desc ibed39.
The sys em comp ised a silicon chip wi h a hicknesses ( ) and wid hs (w) suppo ed by a pai o hea e s
(Fig.2B), esul ing in a hea lux P1 be ween he hea e s h ough he mic o luidic chip as:
λ
=
⋅
⋅ΔP
w
L
T,
(1)
1Si
whe e λSi is he he mal conduc ance o Si, (L) is he chip leng h, and ΔT is he T di e ence be ween he hea e s.
The esul ing alue o P1 was calcula ed o be ≈ 2.42 W.
The con ec ion (P2) in ai was:
=⋅⋅⋅ −PDwh TT(),(2)
21
whe e D is he dis ance be ween he hea e s, h is he con ec ion coe icien and T1 ambien empe a u e, gi -
ing an ampli ude o P2 as ≈ 259 mW, which is 10.8% esul ing in mino non-uni o mi y o he ∇T along he
mic ochannel.
The powe dissipa ed (P3) due o a wa e -based sample (sample) ν in luence can be calcula ed by:
ρ=⋅υ⋅ ⋅Δ·PcT,(3)
31
whe e ρ is he speci ic mass o he sample and c is he sample hea capaci ance.
Finally, he e is also a adia ion powe (P4) emi ed in luencing he ∇T alue. The P4 ampli ude is de ined by
he S e an-Bol zmann law:
=⋅⋅ε σ⋅ −
·PDwTT
()
,(4)
441
4
whe e ε is su ace emissi i y, and σ is he S e an-Bol zmann cons an . We calcula ed he alues o P1, P2, P3, and
P4 as be o e39 and de e mined he o al alues o (P2 + P3 + P4) a e only 9% o he P1; hus, hei in luence on he
∇T can be neglec ed.
We neglec ed di e en empe a u es along he chip o make he analy ical es ima ion simple . I should be
done mo e complexly, using he in eg al o he unc ion along he g adien . Ne e heless, he ini e elemen mod-
elling (FEM) does ake he local empe a u e in o accoun .
FEM and IR imaging. We modeled he mic o luidic chip as well as simpli ied hea e s in CAD so wa e and
ans e ed hem o he ini e elemen modeling so wa e COMSOL Mul iphysics. Then we modeled empe a u e
dis ibu ion along he mic ochannel due o con ec ion, adia ion, and sample ν using Hea T ans e in Solids and
Fluids and C eeping Flow modules in a ashion simila o be o e39. He e, we only show he model mesh (Fig.3A)
and calcula ed empe a u e dis ibu ion wi hin he chip and he hea e s (Fig.3B). The simpli ied de ail o he chip
edge showing he bu ied mic ochannel is shown in Fig.3C, whe e he ed line shows he loca ion o empe a u e
da a ex ac ion o subsequen analysis. We pe o med FEM o he sys em wi h no ex e nal in luence, wi h con-
ec ion, adia ion, and bo h o ex ac he empe a u e alues inside he channel (Fig.3D). We also showed he
e ec o con ec ion and adia ion in he inse . Finally, we checked he in luence o he sample ν (Fig.3E), shown
in de ail in he inse .
Then we assembled he sys em and imaged he su ace empe a u e using an IR came a a e a aching he ca -
bon ape on he chip su ace o achie e a uni o m emissi i y close o uni y. The chip was moun ed on wo hea e s
wi h he empe a u e se a ≈ 50 °C and ≈ 80 °C in o de o c ea e he ∇T along he mic ochannel inside he chip.
The i s se o IR images was cap u ed wi hou ν o he wa e sample (Fig.4A), and he second se o images was
aken wi h he ν se o 0.5 µL·s−1 (Fig.4B), all in a monoch oma ic ashion. He e, we used alse colo s o con as
5
Scien i ic RepoR S | (2020) 10:6925 | h ps://doi.o g/10.1038/s41598-020-63620-5
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
enhancemen . The empe a u e p o ile alongside he cen e mic ochannel—indica ed by he black line in Figs.4A
and 4B—was ex ac ed om he IR images and plo ed in Fig.4C. We ound ha he alue o ∇T be ween he wo
hea e s was cons an , con i ming he negligible in luence o con ec ion and adia ion on hea losses as simula ed
by FEM. The ν alues o up o 0.5 µL·s−1 had only a ma ginal di e ence on he se up sys em shi ing he T ampli-
ude by (−2.8 ± 4.5) 0.001·A.U. (mean ± measu emen e o ) (Fig.4C inse ). This empe a u e shi does show
he induced in luence, albei only o a ma ginal alue wi h a ela i ely la ge i ing e o .
M de e mina ion by a low- h ough sys em. P e iously, we showed a me hod o p ecise TM de e mi-
na ion, empe a u e calib a ion, o hea ans e a e measu emen using dsDNA in he p esence o a luo escen
in e cala o such as SYBR G een I o E a G een o s a iona y d ople s40, as well as a low- h ough sys em39. Fo
he la e , we used a mic oscope o cap u e and analyze luo escence images om he mic ochannel ha ing ∇T.
He e, we eplaced a bulky mic oscope wi h a he mally egula ed op o luidic pla o m o cha ac e ize he p o-
posed de ec ion sys em in a bulk sample wi hou ∇T. Fi s , we illed he mic ochannel wi h a dsDNA sample. We
se T1 = T2 and g adually inc eased hei empe a u e alues, s a ing om ≈ 30 °C o ≈ 85 °C wi h ≈ 5 K inc e-
men s while moni o ing he VL ampli ude (Fig.5A). The measu emen a a de e mined empe a u e was ollowed
by washing he solu ion ha in e ac ed wi h lase ligh wi h a esh solu ion; hen, we wai ed o ≈ 10 s o he VL
signal o s abilize he pho obleaching e ec . The VL alue o each empe a u e was ead when he empe a u e
changed on bo h hea e s, and he signal d opped and s abilized as indica ed by he a ows in Fig.5A. Then we
plo ed he VL alues as a unc ion o empe a u e and pe o med a nonlinea cu e i ing using he sigmoidal
Bol zmann unc ion (Fig.5B black line). Figu e5B also shows nega i e nume ical de i a i e o VL wi h espec o
empe a u e (Fig.5B blue line). The maximum o his cu e is he alue o TM as (59.9 ± 0.2)°C (mean ± i ing
e o ), which was close o he TM alue measu ed ea lie using he comme cial qPCR sys em.
M and hea ans e de e mina ion using segmen ed low. Fu he , we conduc ed an expe imen
using segmen ed low o demons a e a apid MCA o he dsDNA a ∇T. The segmen ed low was gene a ed
Figu e 3. FEM o he mic o luidic chip. (A) Mesh in he COMSOL Mul iphysics. (B) The empe a u e
dis ibu ion simula ed by COMSOL wi h he hea e s’ empe a u e alue se o 50 °C and 80 °C. (C) A simpli ied
ske ch showing he edge o he chip wi h a ed line in he bu ied mic ochannel cen e om which he
empe a u e alues we e ex ac ed. (D) Dis ibu ion o empe a u e alues along he mic ochannel (black line)
conside ing he in luence o con ec ion ( ed line), adia ion (blue line), and bo h (g een line) a ze o ν. The inse
is he empe a u e di e ence be ween models conside ing con ec ion and/o adia ing and he basic one using
colo ma king co esponding o he main g aph. (E) Tempe a u e dis ibu ion along he mic ochannel wi h ν
alue as pa ame e . (inse ) Di e ence be ween models wi h di e en ν alues and he one wi h ν = 0 µL·s−1.

6
Scien i ic RepoR S | (2020) 10:6925 | h ps://doi.o g/10.1038/s41598-020-63620-5
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
Figu e 4. In a ed image o ∇T on a chip (A) wi hou liquid low and (B) wi h he ν o ≈0.5 µL·s-1. (C) The
ex ac ed empe a u e as a unc ion o L wi h he slope ep esen ing he ∇T. Black line and ed line ep esen ν
o 0 µL·s-1 and ≈0.5 µL·s-1, espec i ely. Inse : he plo o di e ence in ex ac ed alues wi h and wi hou ν.
Figu e 5. (A) VL ampli ude as a unc ion o ime wi h he empe a u e o bo h hea e s se o alues in he ange
om ≈ 30 °C o ≈ 85 °C using dsDNA wi h TM alue o ≈ 60.6 °C wi h s agnan sample. Each a ow indica es
he ins ance when he VL alue was ead o he s a ed empe a u e. (B) Nonlinea cu e i ing o VL alues as
a unc ion o empe a u e ex ac ed om (A) (black squa es) using he Bol zmann (sigmoid) cu e (black line
wi h squa es) and i s de i a i e (blue line) o de e mine he alue o TM. The e o ba s co espond o s anda d
de ia ions (SD) om h ee measu emen s.
7
Scien i ic RepoR S | (2020) 10:6925 | h ps://doi.o g/10.1038/s41598-020-63620-5
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
o -chip using a double T-junc ion, as i was ound ha he uni o mi y o p oduced segmen s was mo e s able
han one gene a ed wi hin a hea ed chip.
We con olled he d ople gene a ion as well as he a io be ween oil and wa e phases using ex e nal p alues
se om ≈ 25 kPa o ≈ 75 kPa, espec i ely. We eco ded he VL signal o ≈ 250 s, ge ing i s alue o al e na e
be ween ≈ 300 mV and ≈ 3 V o e e ence d ople s and he sample, espec i ely (Fig.6A), wi h he leng h o he
d ople es ima ed o be ≈ 200 µm (Fig.6A inse ).
The VL ampli ude inc eases abo e he baseline co esponding o he d ople en e ing he op o luidic pa h,
eaching he maximum when he whole d ople is inside he op o luidic mic ochannel (Fig.6B). The d ople
he e was exposed o ∇T inside he mic ochannel; hus, he alue o VL d opped. This dec ease o he VL alue is
nonlinea due o he na u e o dsDNA’s mel ing cha ac e is ic. We ex ac ed he pa o he cu e co esponding
o ∇T exposu e ( ed a owed line) and pe o med a non-linea cu e i ing using he Bol zmann (sigmoid)
unc ion wi h espec o ime, subsequen ly con e ing ime o loca ion and empe a u e39. The i ed cu e was
p ocessed by de i a ion -dV/dT (Fig.6C) wi h low speed ( ) alues (Table1) as a pa ame e . The alues o we e
de e mined by he ime ( ) equi ed o a d ople o pass he whole leng h o ≈ 24 mm o mic o luidic mic ochan-
nel and hen = /24. Subsequen ly, we exp essed he TM as a unc ion o alues (Table1 and Fig.6D). Figu e6D
shows how he TM alues inc ease wi h inc easing co esponding o he hea ans e a e. F om his cu e, he
Figu e 6. (A) Fluo escen emission signal om dsDNA sample and e e ence d ople s mo ing ac oss ∇T
on he chip o med by se ing he empe a u e o hea e s a alues o 50 °C and 80 °C wi h hei dis ance o ≈
13 mm. Inse shows he gene a ed luo escence d ople ≈ 200 µm in he mic ochannel. (B) In e p e a ion o
luo escence signal om a d ople passing h ough he ∇T zone in he op ical mic ochannel wi h he alue o
≈ 1.65 mm·s−1. (C) De i a ion o luo escence ampli ude gene a ed om a d ople as a unc ion o . (D) Plo o
TM alues as a unc ion o .
Se p (kPa) ν (µL·s−1) (s) (mm·s−1)TM (°C) (mean ± SD)
≈25 0.22 ≈96.0 ≈0.25 60.13 ± 0.03
≈40 0.28 ≈40.2 ≈0.59 60.64 ± 0.08
≈50 0.33 ≈20.9 ≈1.15 61.19 ± 0.08
≈60 0.37 ≈17.0 ≈1.41 61.55 ± 0.04
≈75 0.45 ≈14.5 ≈1.65 61.97 ± 0.05
Table 1. Mel ing empe a u e as unc ion o low a es in d ople -based con igu a ion, showing he du a ion o
MCA as he alue o ime ( ) equi ed o he d ople o pass hough he mic ochannel.
8
Scien i ic RepoR S | (2020) 10:6925 | h ps://doi.o g/10.1038/s41598-020-63620-5
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
slope can be ex ac ed o calcula e he ime equi ed o hea o each he cen e o he mic ochannel om he
channel sidewalls.
Thus, we pe o med linea i ing o ob ain he slope (S) as (1.28 ± 0.04) K·s·mm−1 (mean ± i ing e o ). The
S alue di ided by ∇T o ≈ 2.31 K mm−1 p oduces a hea ans e ime o ≈ 554 ms. Knowledge o his alue can
be used o co ec he sys em esul s when hey a e collec ed a di e en alues o .
The MCAs a e ypically pe o med using comme cial eal- ime PCR sys ems wi h a ypical empe a u e
amp a e be ween ≈ 0.1 and ≈ 0.5 K·s−1, hus co esponding o 500 s o 100 s pe es (50 K empe a u e amp-
ing), excluding sample loading41. An ul a as me hod o pe o m he MCA in an as onishing ≈ 50 ms has been
epo ed42. I was based on a d ople placed on a cold Cu subs a e wi h he sample hea ed by lase i adia ion
ha equi ed a complex se up and ca e ul sample p epa a ion. Ou op o luidic pla o m enables apid MCA o
he sample con aining biomolecules such as DNA o p o eins. The ime needed o he analysis is less han ≈
30 s, which makes his pla o m as e han o he sys ems. Tha can signi ican ly speed up he MCA and make i
sui able o high- h oughpu sc eening.
conclusion
We p oposed an op o luidic chip ha ing ∇T along he mic ochannel o pe o m he MCA o dsDNA o p o ein
un olding. We i s nume ically analyzed he in luence o con ec ion, adia ion, and a sample ν on he empe a-
u e dis ibu ion, as well as ∇T de ia ion om a cons an alue. The op o luidic chip was equipped wi h a bu ied
mic ochannel wi h a diame e o ≈ 500 µm. We inse ed bo h op ical ibe s and mic ocapilla ies inside he chip,
o ming an in eg a ed and obus au onomous sys em. The mic o luidic channel se ed as a ligh guide o sen-
si i e luo escence de ec ion along he ∇T inside he chip. We demons a ed his concep by de e mining he
dsDNA TM in a con inuous- low con igu a ion. We also conduc ed he MCA in a d ople -based con igu a ion
using segmen ed low. This concep o luo escence moni o ing in a mic ochannel exposed o ∇T ep esen s
a as and cos -e ec i e app oach o he cha ac e iza ion o he mal p ope ies o biomolecules. I could be
c ucial in many a eas o biology and chemis y, including he s abili y o p o eins, by exposing hem o di e en
en i onmen s such as pH, bu e composi ion and ionic s eng h, pu i y con ol, and p o ein-ligand in e ac ion.
Those applica ions could ha e a g ea impac on d ug disco e ies as well as he molecula diagnos ics o in ec ious
diseases o geno yping.
Recei ed: 26 No embe 2019; Accep ed: 30 Ma ch 2020;
Published: xx xx xxxx
Re e ences
1. Dolan, E. B., Haugh, M. G., Tallon, D., Casey, C. & McNama a, L. M. Hea -shock-induced cellula esponses o empe a u e
ele a ions occu ing du ing o hopaedic cu ing. J R Soc 9, 3503–3513, h ps://doi.o g/10.1098/ si .2012.0520 (2012).
2. Ve ga a, M. e al. Di e en ial E ec o Cul u e Tempe a u e and Speci ic G ow h Ra e on CHO Cell Beha io in Chemos a Cul u e.
PLOS ONE 9, e93865, h ps://doi.o g/10.1371/jou nal.pone.0093865 (2014).
3. Ka a a, M. e al. C i ical s uc u al luc ua ions o p o eins upon he mal un olding challenge he Lindemann c i e ion. P oc. Na l.
Acad. Sci. USA, 201707357, h ps://doi.o g/10.1073/pnas.1707357114 (2017).
4. Tanaka, J. e al. KRAS geno yping by digi al PCR combined wi h mel ing cu e analysis. Sci Rep 9, 2626, h ps://doi.o g/10.1038/
s41598-019-38822-1 (2019).
5. Mi alles, V., Hue e, A., Malloggi, F. & Jullien, M.-C. A Re iew o Hea ing and Tempe a u e Con ol in Mic o luidic Sys ems:
Techniques and Applica ions. Diagnos ics 3, 33–67, h ps://doi.o g/10.3390/diagnos ics3010033 (2013).
6. Sohel Mu shed, S. M., Tan, S.-H. & Nguyen, N.-T. Tempe a u e dependence o in e acial p ope ies and iscosi y o nano luids o
d ople -based mic o luidics. J. Phys. D 41, 085502, h ps://doi.o g/10.1088/0022-3727/41/8/085502 (2008).
7. Mu shed, S. M. S., Tan, S. H., Nguyen, N. T., Wong, T. N. & Yobas, L. Mic od ople o ma ion o wa e and nano luids in hea -
induced mic o luidic T-junc ion. Mic o luid Nano luid 6, 253–259, h ps://doi.o g/10.1007/s10404-008-0323-3 (2009).
8. Ve neuil, E., Co de o, M., Gallai e, F. & Ba oud, C. N. Lase -Induced Fo ce on a Mic o luidic D op: O igin and Magni ude.
Langmui 25, 5127–5134, h ps://doi.o g/10.1021/la8041605 (2009).
9. Yesiloz, G., Boybay, M. S. & Ren, C. L. E ec i e The mo-Capilla y Mixing in D ople Mic o luidics In eg a ed wi h a Mic owa e
Hea e . Anal Chem 89, 1978–1984, h ps://doi.o g/10.1021/acs.analchem.6b04520 (2017).
10. Kha e , A., Mohammadi, M., Mohamad, A. & Nezhad, A. S. Dynamics o empe a u e-ac ua ed d ople s wi hin mic o luidics. Sci
Rep 9, 3832, h ps://doi.o g/10.1038/s41598-019-40069-9 (2019).
11. Ho, P. C. & Nguyen, N.-T. Nume ical s udy o he mocoalescence o mic od ople s in a mic o luidic chambe . Phys. Fluids 25,
082006, h ps://doi.o g/10.1063/1.4819134 (2013).
12. Liu, J.-H., Hsia, K.-C., Yokokawa, R. & Lu, Y.-W. Mic o ubule polyme iza ion in alignmen by an on-chip empe a u e g adien
pla o m. Sens. Ac ua o B-Chem. 298, 126813, h ps://doi.o g/10.1016/j.snb.2019.126813 (2019).
13. Li, K.-C., Ding, S.-T., Lin, E.-C., Wang, L. A. & Lu, Y.-W. Mel ing analysis on mic obeads in apid empe a u e-g adien inside
mic ochannels o single nucleo ide polymo phisms de ec ion. Biomic o luidics 8, 064109–064109, h ps://doi.
o g/10.1063/1.4902907 (2014).
14. Dodge, A., Tu ca i, G., Law ence, I., de Rooij, N. F. & Ve poo e, E. A Mic o luidic Pla o m Using Molecula Beacon-Based
Tempe a u e Calib a ion o The mal Dehyb idiza ion o Su ace-Bound DNA. Anal Chem 76, 1778–1787, h ps://doi.o g/10.1021/
ac034377+ (2004).
15. Neuzil, P., Pippe , J. & Hsieh, T. M. Disposable eal- ime mic oPCR de ice: lab-on-a-chip a a low cos . Mol. Biosys . 2, 292–298,
h ps://doi.o g/10.1039/B605957K (2006).
16. Da ies, E., Ch is odoulides, P., Flo ides, G. & Kalli, K. Mic o luidic Flows and Hea T ans e and Thei In luence on Op ical Modes
in Mic os uc u e Fibe s. Ma e ials 7, 7566–7582, h ps://doi.o g/10.3390/ma7117566 (2014).
17. Toh, K. C., Chen, X. Y. & Chai, J. C. Nume ical compu a ion o luid low and hea ans e in mic ochannels. In . J. Hea Mass T ans
45, 5133–5141, h ps://doi.o g/10.1016/S0017-9310(02)00223-5 (2002).
18. Hajmohammadi, M. R., Alipou , P. & Pa sa, H. Mic o luidic e ec s on he hea ans e enhancemen and op imal design o
mic ochannels hea sinks. In . J. Hea Mass T ans 126, 808–815, h ps://doi.o g/10.1016/j.ijhea mass ans e .2018.06.037 (2018).
19. No ak, L., Neuzil, P., Pippe , J., Zhang, Y. & Lee, S. An in eg a ed luo escence de ec ion sys em o lab-on-a-chip applica ions. Lab
Chip 7, 27–29 (2007).
20. Ah be g, C. D., Ilic, B. R., Manz, A. & Neužil, P. Handheld eal- ime PCR de ice. Lab Chip 16, 586–592 (2016).
21. Ah be g, C. D., Manz, A. & Neužil, P. Palm-sized de ice o poin -o -ca e Ebola de ec ion. Anal Chem 88, 4803–4807 (2016).
9
Scien i ic RepoR S | (2020) 10:6925 | h ps://doi.o g/10.1038/s41598-020-63620-5
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
22. Neuzil, P. e al. Rapid de ec ion o i al RNA by a pocke -size eal- ime PCR sys em. Lab Chip 10, 2632–2634 (2010).
23. Ozcelik, D., Cai, H., Leake, K. D., Hawkins, A. R. & Schmid , H. Op o luidic bioanalysis: undamen als and applica ions.
Nanopho onics 6, 647–661, h ps://doi.o g/10.1515/nanoph-2016-0156 (2017).
24. Fan, X. & Whi e, I. M. Op o luidic mic osys ems o chemical and biological analysis. Na Pho onics 5, 591–597, h ps://doi.
o g/10.1038/npho on.2011.206 (2011).
25. Song, C. & Tan, S. H. A Pe spec i e on he Rise o Op o luidics and he Fu u e. Mic omachines (Basel) 8, 152, h ps://doi.o g/10.3390/
mi8050152 (2017).
26. Pe siche i, G., G imaldi, I. A., Tes a, G. & Be nini, R. Mul i unc ional op o luidic lab-on-chip pla o m o Raman and luo escence
spec oscopic mic o luidic analysis. Lab Chip 17, 2631–2639, h ps://doi.o g/10.1039/C7LC00460E (2017).
27. Yang, T., S a akis, S. & deMello, A. A High-Sensi i i y, In eg a ed Abso bance and Fluo escence De ec ion Scheme o P obing
Picoli e -Volume D ople s in Segmen ed Flows. Anal Chem 89, 12880–12887, h ps://doi.o g/10.1021/acs.analchem.7b03526
(2017).
28. Cai, H. e al. Op o luidic analysis sys em o ampli ica ion- ee, di ec de ec ion o Ebola in ec ion. Sci Rep 5, 14494–14494, h ps://
doi.o g/10.1038/s ep14494 (2015).
29. Sche , S. M. e al. Real-Time Cap u e and Visualiza ion o Indi idual Vi uses in Complex Media. ACS Nano 10, 2827–2833, h ps://
doi.o g/10.1021/acsnano.5b07948 (2016).
30. Zhu, H., Ma andadi, S., Coskun, A. F., Yaglide e, O. & Ozcan, A. Op o luidic luo escen imaging cy ome y on a cell phone. Anal
Chem 83, 6641–6647, h ps://doi.o g/10.1021/ac201587a (2011).
31. Bal am, K. C. e al. The Nanoli hog aphy Toolbox. J. Res. Na l. Ins . S and. 121, 464–475, h ps://doi.o g/10.6028/j es.121.024 (2016).
32. Gablech, I. e al. Fab ica ion o bu ied mic o luidic channels wi h obse a ion windows using em osecond lase pho oabla ion and
pa ylene-C coa ing. Mic o luid Nano luid 22, 105 (2018).
33. Ilic, B., Czaplewski, D., Zalalu dino , M., Schmid , B. & C aighead, H. Fab ica ion o lexible polyme ubes o mic o and
nano luidic applica ions. Jou nal o Vacuum Science & Technology B: Mic oelec onics and Nanome e S uc u es P ocessing,
Measu emen , and Phenomena 20, 2459–2465 (2002).
34. Lae me, F., Schilp, A., Funk, K. & O enbe g, M. In Technical Diges . IEEE In e na ional MEMS 99 Con e ence. Twel h IEEE
In e na ional Con e ence on Mic o Elec o Mechanical Sys ems (Ca . No. 99CH36291). 211–216 (IEEE).
35. Reismann, M., B e schneide , J. C., Plessen, G. V. & Simon, U. Re e sible Pho o he mal Mel ing o DNA in DNA–Gold-Nanopa icle
Ne wo ks. Small 4, 607–610, h ps://doi.o g/10.1002/smll.200701317 (2008).
36. Cas o, E. R. e al. De e mina ion o dynamic con ac angles wi hin mic o luidic de ices. Mic o luid Nano luid 22, 51 (2018).
37. S a os, V., Gablech, I., Ilic, B. R., Peka ek, J. & Neuzil, P. In si u obse a ion o ca bon nano ube laye g ow h on mic obolome e s
wi h subs a es a ambien empe a u e. J. Appl. Phys 123, 7, h ps://doi.o g/10.1063/1.5016465 (2018).
38. S a oš, V., Gablech, I., Peká ek, J., Klempa, J. & Neužil, P. P ecise de e mina ion o he mal pa ame e s o a mic obolome e . In a ed
Phys Techn 93, 286–290 (2018).
39. Zhu, H. e al. Hea ans e ime de e mina ion based on DNA mel ing cu e analysis. Mic o luidics and Nano luidics 24, 7 (2020).
40. Neuzil, P., Cheng, F., Soon, J. B. W., Qian, L. L. & Reboud, J. Non-con ac luo escen bleaching-independen me hod o empe a u e
measu emen in mic o luidic sys ems based on DNA mel ing cu es. Lab Chip 10, 2818–2821, h ps://doi.o g/10.1039/c005243d
(2010).
41. Ba sch, M. S. e al. The o a y zone he mal cycle : a low-powe sys em enabling au oma ed apid PCR. PloS one 10,
e0118182–e0118182, h ps://doi.o g/10.1371/jou nal.pone.0118182 (2015).
42. Baaske, P., Duh , S. & B aun, D. Mel ing cu e analysis in a snapsho . Applied Physics Le e s 91, 133901 (2007).
Acknowledgemen s
The au ho s wish o acknowledge he inancial suppo o he W099109 g an om P.R. China, he GACR P ojec
Numbe GA16-11140S om he Czech Republic, and he echnical suppo p o ided by Nano+ and by Cen um
SIX o B no Uni e si y o Technology, Czech Republic. LY and SN acknowledge he unding om he Resea ch
G an Council o Hong Kong, ecei ed unde G an Numbe 16209316. The au ho s would like o hank Pe e
Fecko and Jakub Somme o hei help wi h chip ab ica ion.
Au ho con ibu ions
Z.F. pe o med analysis, in ep e a ion o da a and w o e he main manusc ip ex H.Z. was esponsible o chip
design and ab ica ion J.H. e iew & edi ing S.N. and L.Y. we e esponsible o FEM simula ions and pa icipa ion
on w i ing he manusc ip P.P. and A.O. we e esponsible o op o luidic se up and measu emen s P.N. w i ing –
e iew & edi ing All au ho s e iewed he manusc ip .
compe ing in e es s
The au ho s decla e no compe ing in e es s.
Addi ional in o ma ion
Co espondence and eques s o ma e ials should be add essed o P.N.
Rep in s and pe missions in o ma ion is a ailable a www.na u e.com/ ep in s.
Publishe ’s no e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and
ins i u ional a ilia ions.
Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional
License, which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o
o ma , as long as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C e-
a i e Commons license, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his
a icle a e included in he a icle’s C ea i e Commons license, unless indica ed o he wise in a c edi line o he
ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons license and you in ended use is no pe -
mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly om he
copy igh holde . To iew a copy o his license, isi h p://c ea i ecommons.o g/licenses/by/4.0/.
© The Au ho (s) 2020