Single File Flow of Biomimetic Beads for Continuous SERS ...downloads.hindawi.com/journals/acmp/2018/2849175.pdf · AdvancesinCondensedMatterPhysics 10 9 8 7 6 5 4 3 2 1 0 Speed (mm/s)

Research ArticleSingle File Flow of Biomimetic Beads for Continuous SERSRecording in a Microfluidic Device

Diego Calzavara12 Davide Ferraro1 Lucio Litti2 Greshia Cappozzo12 Giampaolo Mistura1

MorenoMeneghetti 2 andMatteo Pierno 1

1Dipartimento di Fisica e Astronomia ldquoGalileo Galileirdquo Universita di Padova Via Marzolo 8 35131 Padova Italy2Dipartimento di Scienze Chimiche Universita di Padova Via Marzolo 1 35131 Padova Italy

Correspondence should be addressed to Moreno Meneghetti morenomeneghettiunipditand Matteo Pierno matteopiernounipdit

Received 31 January 2018 Revised 3 May 2018 Accepted 6 May 2018 Published 20 June 2018

Academic Editor Mohindar S Seehra

Copyright copy 2018 DiegoCalzavara et alThis is an open access article distributed under the Creative CommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

A major challenge in cancer treatment is the quantification of biomarkers associated with a specific cancer type Importantbiomarkers are the circulating tumor cells (CTCs) detached from the main cancer and circulating in the blood CTCs are veryrare and their identification is still an issue Although CTCs quantification can be estimated by using fluorescent markers all thefluorescence techniques are strongly limited by the number of emissions (therefore markers) that can be discriminated with oneexciting line by their bleaching characteristics and by the intrinsic autofluorescence of biological samples An emerging techniquethat can overcome these limitations is Surface Enhanced Raman Scattering (SERS) Signals of vibrational origin with intensitysimilar to those of fluorescence but narrower bandwidths can be easily discriminated even by exciting with a single laser line Werecently showed the benefit of this method with cells fixed on a surface However this approach is too demanding to be applied inclinical routine To effectively increase the throughput of the SERS analysismicrofluidics represents a promising toolWe report twodifferent hydrodynamic strategies based on device geometry and liquids viscosity to successfully combine a microfluidic designwith SERS

1 Introduction

One of the most common techniques to characterize cellsis cytofluorometry which in a continuous flow allowsdetecting cells labeled with specific fluorescent probes Upto thousands of cells per second can be processed How-ever despite its rapidity this technique is still limited tofew detection channels because of the characteristic largebandwidths of fluorescence emissions and the use of dif-ferent exciting lines for different fluorescent probes In thisscenario Surface Enhanced Raman Scattering (SERS) [1]is a promising alternative because it shows intense signalssimilar to those of fluorescence but also very sharp bandsbeing of vibrational origin This characteristic allows an easymultiplexing approach with many detected signals excitedwith only one laser line [1] Furthermore bleaching whichis a problem for fluorescence is not present for SERS andautofluorescence of biological samples does not disturb SERS

signals The enhancement of Raman scattering derives fromthe amplification of the local electromagnetic fields on thesurface of plasmonic nanostructures Optimized gold nanos-tructures functionalized for cell targeting were found tobe good plasmonic nanostructures useful to detect differentpopulations of cancer cells [2 3] However cells were fixedon a surface and analyzed one by one under a micro-Ramanmicroscope limiting the time analysis and accordingly thethroughput of the system

Microfluidics which consists in the manipulation ofliquids inside microchannels [4] due also to the simplicityof the typical microfabrication strategies [5ndash7] represents apromising platform to overcome this limitation especiallyrelated to chemical [8 9] and biomedical applications [1011] As a matter of fact during the past years microfluidicsshowed capabilities in performing several biomedical appli-cations regarding nucleic acid [12ndash15] immunoassays [16ndash19]and cell manipulations [20 21]

HindawiAdvances in Condensed Matter PhysicsVolume 2018 Article ID 2849175 9 pageshttpsdoiorg10115520182849175

2 Advances in Condensed Matter Physics

Examples of microfluidic devices coupled with Ramanspectroscopy have been successfully used for sample precon-centration or storing cells during the spectroscopic analysis[22ndash24]

For example dielectrophoresis (DIE) is largely used formanipulating particles in microfluidic devices [25 26] andmore recently this approach has been recently combinedwithSERS analysis [27] for detection of specific biomarkers [28]and cells [29] However although DIE offers the advantageof precisely positioning cells or particles in the channel itrequires devices with integrated electrodes showing specificgeometry increasing the complexity and the cost of themicrofabrication processes and preventing the productionof disposable device [30] A passive approach that requiressimpler fabrication protocols is represented by the use of aflow-focusing geometry to align cells (or particles) in a singleline and force them to pass under the Raman laser spotssimilarly to what is shown with UV-visible spectroscopy inflow [31]However only onework has been presented so far topush forward the flow-focusing geometry with SERS analysis[32] In fact while this device allowed acquiring continuouslythe signal from flowing cells it presents twomain drawbacksfirst the usability of this procedure is severely limited bythe acquisition time of SERS spectrometers (given by theCCD exposure and shutter) In addition even with fastspectrometers (100-150ms of acquisition time) it is necessaryto design themicrofluidic devices to reduce the velocity of thecarrying flow Since this achievement is commonly obtainedby reenlarging the carrying flow after single file confinement[33] it has been demonstrated that beads flowing in amicrofluidic device can be laterally displaced according totheir size and stiffness [34 35] This may introduce someuncontrolled deviation from the laser spot an issue whendealing with clinical analysis

Given that the fluidic manipulation of cells (or beads) atslow speed in amicrofluidic device is highly demanding sinceit requiresworking at very lowpressure (typically hundreds ofPa) causing nonstable flows a different approach can exploitthe viscosity 120583 of the carrier fluid solution

Within this approach it is possible to affect the hydraulicresistance 119877ℎ of the microfluidic device and so the final flowrate As a matter of fact in a microchannel once fixed thepressure drop Δ119875 at the channel ends and the flow rate 120593 canbe calculated as [36]

120593 =Δ119875

119877ℎ(1)

119877ℎ for a microfluidic channel of length 119871 and radius 119903 can beapproximated as [37]

119877ℎ =8120583119871

1205871199034(2)

Therefore with fixed geometry an increase of 120583 leads to aproportional increase of119877ℎ and thus a reduction of the globalflow rate in the device (see (1))

We believe that both strategies for reducing the cells speedare interesting and both show advantages and disadvantages(i) changing the geometry is very simple in the fabrication

step however it leads to a nonperfect straight trajectory ofthe cells (ii) acting on the viscosity of the liquid could be notalways possible however it permits working with a narrowedconfined flow and thus a deterministic path of the cells Inthis paper we will compare the two strategies demonstratinghow even the latter one can be exploited to perform SERSin a microfluidic device We used for proof-of-conceptexperiments biotinylated polystyrene beads (PS beads) ascell model [38 39] SERS nanostructures produced withgold nanoparticles functionalized with an optimized SERSreporter (NPCy(SH)2) were also functionalized with strep-tavidin The biotin presented on the PS beads and the strep-tavidin on the gold nanostructures represent the antigen-antibody pair used for the identification of cancer cells

First we introduce the methods for the fabrication ofthe microfluidic devices the production of the SERS activenanostructures and the experimental setup Then we reportthe numerical and experimental results for the fluidic charac-terization and we demonstrate the possibility of performingSERS analysis of beads flowing in the microfluidic deviceFinally we conclude comparing the two fluidic strategies andthe results

2 Materials and Methods

21 Microfluidic Device Fabrication and LiquidHandling Themicrofluidic device was produced by conventional replicamolding [5 40] with polydimethylsiloxane (or PDMS Syl-gard 184 by Dow Corning) of a mold prepared by pho-tolithography with SU-8 (by MicroChem) In detail afterhaving produced the mold the nonpolymerized PDMS waspoured on it and placed in an oven at 70∘C for 1h Afterthat the PDMS part was peeled off punched for inlets andbonded to a glass slide by Oxygen Plasma treatment FinallyPTFE tubing having 03mm (06mm) inner (outer) diameter(by Sigma Aldrich) were fixed by epoxy glue as shown inFigure 1(a) The final design described in Figure 1(b) showsa flow-focusing junction aimed at creating a focused flow inthe center channel The device presents two inletsrsquo channelone for the carrier liquid used to generate the confined flowand the other for the liquids containing the functionalizedbeads both have the same cross section 100120583m times20120583m oflateral and thickness dimensions respectivelyThen after thejunction the narrow channel opened in a larger chamber(w=700120583m) aimed at reducing the flow rate and thus thespeed of the beads However as introduced before the con-fined stream flow in the chamber is larger than in the channel(see Figure 1(b)) and thus the beads position becomes lessdetermined (see also Movie S1 in Supplementary Materials)

In order to optimize the focusing condition and charac-terize the speed of the beads different glycerolwater mixeswere used as carrier fluid varying the viscosity between 1 and120 cP then during the flow characterization experimentsthe inner fluid was composed of non-coated polystyrene(PS) beads (size 80-129120583m by Spherotech) dispersed in aPercollwater solution (by GE Healthcare) at 42 ww Thelatter was used to match the density of the PS beads in orderto reduce the sedimentation

Advances in Condensed Matter Physics 3

PDMS

glass slide

1 23

PTFE tubing

(a)

100m

Flow Floww

beads

carrier fluid

inner fluid

D

０1

０2０3

＄

(b)

Figure 1 (a) Picture of the microfluidic device presenting inlets (12) and outlet (3) (b) Scheme of the microfluidic device the carrier liquid(light blue) confines the inner fluid (green) that contains the PS beads (red spots)The scheme shows that the focusing is more efficient in thechannel region (D) than in the chamber part (D1015840) P1 P2 and P3 are the pressure corresponding to the inletsoutlet shown in (a)

The liquids in the microfluidic device were manipulatedby an external pressure controller (MFCS by Fluigent) andthe applied pressure was varied between 01kPa and 25kPain order to find the best condition for the flow confinement(see ldquoDrdquo in Figure 1(b))

22 Polystyrene Beads for SERS Detection Preparation

(i) Streptavidin Functionalization with a Thiol Group Strep-tavidin was thiolated using a procedure already reported [2]Briefly the protein is mixed with 2-iminothiolane (110 ratio)in a 01 M NaHCO3 solution The reaction proceeded for 2hours at room temperature and at 4∘CovernightThemixturewas purified by dialysis and dispersed in PBS

(ii) AuNS Synthesis Gold nanoparticles were synthetized bylaser ablation producing highly pure and stable colloids aspreviously reported [41] In detail a NdYAG (Quantel) laserat 1064 nm with 9 ns pulses and 10 Hz repetition rate wasfocalized on the surface of 9999 pure gold plate at thebottom of a flask filled with 1 120583M NaCl aqueous solution100 120583L of 30 120583MNPCy(SH)2 [42] was added to functionalizeand aggregate the particles (AuNS) A controlled aggregationallows producing hot spots in AuNS where the SERS signal isstrongly enhancedThe unreacted NPCy(SH)2 was separatedby centrifugationThe nanostructures were further function-alizedwith the thiolated streptavidinThemixture was stirredfor 3 hours at room temperature and left at 4∘CovernightThenanostructures were then purified by centrifugation UV-vis-NIR (Cary 5000 Agilent) and 120583-Raman (InVia Renishaw)spectroscopies were used to characterize the nanostructures

(iii) Incubation of AuNS with PS Beads Biotinylated PSbeads (size 80-129120583m by Spherotech) were mixed with thefunctionalized AuNS and left for 4 hours at room temper-ature Then the mixture was centrifuged at 1000 RCF for3 minutes The supernatant containing the unreacted AuNS

was discarded and the centrifuged PSAuNSdispersed in thePercollwater solution as for the flow characterization part

(iv) SERS Measurements of Flowing PSAuNS For the con-tinuous SERSmeasurements the 785 nm laser line at about 30mWwas focused on the middle of the sample channel with a20timesmagnification objective The acquisition was continuousat a repetition rate of 1HzThedataset was baseline subtractedand each spectrum filtered for the intensity at 688 cmminus1 Thedata analysis was obtained with homemade Matlab codes

3 Results and Discussion

31 Optimization of the Flow-Focusing Condition At first weinvestigated the best pressure conditions in order to obtain anarrow flow in the center of the channel As a matter of factthis is a fundamental aspect to obtain an aligned set of beadspassing one by one in a specific detection region Figure 2describes the lateral size of the focused flow as a functionof the ratio between the applied pressure P2P1 keepingfixed P3 at 05kPa In order to perform this experiment afluoresceinPercoll solution was used in the inner channeland observed by a fluorescent microscope It is possible tonotice that the trend is almost linear in the investigated regionand in particular for values below 05 D is smaller than 20120583m which corresponds to twice the beads size Thereforein order to ensure stable pressure control and a narrow flowwe fixed their values at P1=25kPa P2=10kPa and P3=05kPa(11987521198751=04) Finally those values were kept constant in all thefollowing experiments in order to evaluate only the influenceof the viscosity on the beadsrsquo speed

32 Size of the Microfluidic Chamber The influence on thelocal speed of the lateral size (w) of the microfluidic chamberwas first evaluated keeping the applied pressure fixed (seeFigure 3(a)) We performed different numerical simulation(with Comsol Multiphysics) varying w and measuring the


(a)

120

100

80

60

40

20

0

D [

m]

00 02 04 06 08 10 12 14

０2０1

(b)

Figure 2 (a) Fluorescent image of the flow-focusing with the fluorescein mix used as inner solution (b) Dimension of D as function of thepressure ratio P2P1 the red circle underlines the selected region for the following experiment

distance (mm)

w

local speed(ms)

dist

ance

(mm

)

r=100m

200m

(a)

15

14

13

12

11

10

09

08

07

06

05

Nor

mal

ized

spee

d (v

Ｐ )

0 100 200 300 400 500 600 700 800 900 1000

w (m)

(b)

Figure 3 (a) 2D surface plots results of the local speed in the used microfluidic channels for a particular value of w (b) Normalized localspeed of the liquid considered 200120583m from the end of the channel and in the center of the chamber plotted as function of its lateral dimensionw The reported simulation results correspond to the red data close to them The local speed (v) was normalized by the speed obtained forthe condition of a straight channel (v1015840) without the chamber (w=100120583m)

droplet speed at 200 120583m from the main channel after thefocusing (vertical dashed line in Figure 3(a)) Figure 3(b)describes the droplet speed of the beads flowing in thecenter of the device as a function of the lateral size of thefluidic chamber All the speeds are normalized by the speedevaluated for a straight channel (w=119903=100120583m) It is possibleto notice that by increasing w the speed initially increasesup to 14 and then decreases reaching a plateau at about 07and it is reduced to 13 of the initial value This behavior canbe explained by considering that working at a fixed pressurethe increase ofw leads to two contributions (i) a reduction ofthe hydraulic resistance of the device that brings an increaseof the flow rate and thus of the speed (ii) an increase ofthe section of the device that causes a local speed reduction

Therefore the observed effect is a balance between these twocontributions Observing the graph at the beginning theformer contribution is more important while for wgt350120583mthe latter one is dominant Finally over a certain value theincrease of the chamber size does not have any effect In themicrofabrication process we have chosen w=700 120583m for ourdevice in order to be in the plateau region

33 Comparison between the Use of the Chamber and theViscous Liquid As introduced above another strategy toreduce the beads speed in the microfluidic device is toincrease the viscosity 120583 of the carrier fluid Figure 4 shows theexperimental results obtained by varying the viscosity of the


10

9

8

7

6

5

4

3

2

1

0

Spee

d (m

ms

)

0 50 100 150 200 250

viscosity (cP)

visc

osity

(cP)

channelchamber

glycerol conc ()

250

200

150

100

50

0

0 20 40 60 80 100

Figure 4 Speed of a beads evaluated in the center of the microfluidic device in the channel (black data) and in the chamber (red data)plotted as function of the viscosity of the carrier liquids The reported speeds are the averaged results of data for at least 5 beads and the errorbars are the related standard deviation The liquid viscosity was varied by different mix of water and glycerol (see inset)

carrier fluid in the main channel and in the large chamberAt first for low viscosity (120583 ge 1) the ratio between the tworegimes is about 13 (chamberchannel≃39) similar to theone obtained by the Comsol simulation which consideredpure water solutions instead of the beadsPercoll mixtureThen taking into account the effect of the viscosity only inthe channel we can observe that for 120583 ge 25 the speed ofthe beads decreases filling this gap reaching the same valueobtained in the chamber for low viscosity (see dashed line inFigure 4) This result shows that in the case of confined flowonly acting on the viscosity of the carrier fluid it is possibleto reduce the speed of the flowing beads asmuch as changingthe channel geometry Additionally as previously describedthe presented approach allows a better confinement of theflowing beads than using a larger chamber

After that going deeper in the analysis the beads can beadditionally slowed down considering the effect of 120583 in thechamber part of the device This suggests that combining thetwo contributions given by the geometry and the viscosity itis possible to reduce the local speed by almost 10 times

Finally both trends reach a plateau at about 40 cPprobably due to the fact that for higher viscosity the flow ofthe carrier fluid does not change anymore at the fixed pressure(P1=25kPa)

34 SERS Nanostructures and Model for SERS Labeled CellsThe above reported data were obtained with PS beadsnot labeled with AuNS For the detection of SERS signalswithin the microchannel PS beads were coupled with thestreptavidin functionalized SERS-AuNS Coupling PS beadswith AuNSmimicked the interaction between cells and SERSlabel [38 43]

At first gold nanoparticles (AuNP) were produced bylaser ablation synthesis in solution (LASiS) as previously

reported for a wide range of materials [44ndash46] BrieflyNdYAG nanosecond laser pulses were focused on a puregold plate in a micromolar NaCl solution Dithiolated siliconnaphthalocyanine (NPCy(SH)2) was used as SERS reporterto functionalize AuNP due to its resonance with the excitinglaser line at 785 nm The function of the double thiol isto bridge the AuNP in order to achieve the maximumenhancement of the local field Consequently the SERSsignals are strongly enhanced

Finally the AuNS were functionalized with the thio-lated streptavidin using a previously reported protocol forantibodies [2 47] as sketched in Figure 5(a) Figure 5(b)reports the extinction spectrum of the gold nanostructures(AuNS) which shows the good resonance with the 785 nmexciting laser line used for the 120583-Raman measurementsFinally Figure 5(c) shows the strong SERS spectrum of AuNSwith the characteristic signature of NPCy(SH)2 the band at688 cmminus1 which will be used as the label of AuNS

PS beads present already a coating of biotin that allowsexploiting the interaction with streptavidin present on thefunctionalized AuNS This allows simulating a typical strongbinding between antigens and antibodies AuNS and PSbeads were mixed together at room temperature for 4 hoursand the unbounded AuNS were discarded by centrifugationThe AuNS-PS beads conjugates were characterized by 120583-Raman and TEM imaging Every PS beads showed theSERS signature of Figure 5(c) TEM images (see Figure 6(a))clearly showed AuNP aggregates on the surface of the PSbeads Finally AuNS-PS beads were dispersed in Percollsolution and injected into the microfluidic device for theSERS detection (see Figure 6(b))

35 Continuous SERS Detection in Microfluidic Device withGlycerol Solution as Carrier Fluid The experiments to record


ThiolatedStreptavidin

AuNPby LASiS

AuNS

０Ｓ(３（)2（2（2３（

（2（2３（

N

NN

N

N

NN

NSi

(a)05

04

03

02

01

00

＜Ｍ＜

=2ＧＧ

300 400 500 600 700 800 900 1000 1100

Wavelength (nm)

(b)

cps

1800

1600

1400

1200

1000

800

600

400

200

1800 20001600140012001000800600400200

0

Raman Shift (＝Ｇ-1)

688 ＝Ｇ-1

(c)

Figure 5 (a) Cartoon representing the synthesis of AuNS from the laser ablation to the functionalizationwith the SERS reporter NPCy(SH2)and streptavidin ((b) and (c)) UV-vis-NIR and SERS spectra of the functionalized AuNS respectively

SERS signals were performed using the AuNS-PS beads Thebeads were flown at the previously characterized pressure(1198751=25kPa 1198752=1kPa and 1198753=05kPa) and focused in thecenter of the channel by the glycerol solution (80 ww)as carrier liquid As shown in Figure 4 in these conditionsthe beads speed was about 3mm sminus1 which was enough tobe detected by the 120583-Raman setup Figure 6(c) describes thesequence of two images of the microfluidic device in whicha PS bead was approaching the sampling area The 785 nmlaser of the 120583-Raman was focused as a line on the centerof the focused flow region just after the focusing junctionUsing a 20timesmagnification the laser spot had a dimension of100 120583m times5 120583m Raman spectra were acquired continuouslywhile the beads were flowing into the channel at a frequencyof 1 spectrum per second The Raman band of PDMS at 710cmminus1(green star in Figure 6(d)) was used to calibrate thefocal position The signal arising from the AuNS is clearlydistinguishable at 688 cmminus1 marked as red star in Figure 6(d)From TEM images as reported in Figure 6(a) one canestimate about 15 of AuNS surface coverage of a PS beadwhich corresponds to about 102 nanostructures per bead Anexample of a continuous flow measurement is presented inFigure 6(e) where the intensity at 688 cmminus1 was monitoredin time and the peaks marked with red stars are due to the

bright SERS signals of AuNS at this specific wavelength Forexample the third peak of Figure 6(e) detected at about 815scorresponds to the spectrum reported in Figure 6(d)

SERS signals of the AuNS are bright signals which can beobtained with engineered nanostructures The results showthat these signals can be seen for the flowing PS beads labeledwith the nanostructures and that they can be easily detectedalso with a conventional 120583-Raman instrument

4 Conclusions

By using a proper microfluidic approach we showed thecapability of achieving continuous SERS analysis by using aflow-focusing device We reported two strategies aimed atslowing down PS beads in order to synchronize their passagetimes with the typical ones used in 120583-Raman technologyWorking either on the device geometry or the liquid viscosityor on both of them we proved a fine tuning of the beadsspeed flowing in a single file at low pressure regime TargetedPS beads used for mimicking targeted cells are clearlyidentified with a continuous flow SERS analysis showing thata microfluidic flow-focusing can be designed for controllinga flow of objects similar to cells with density matched tothe suspending medium and that SERS signals deriving


PS bead

AuNS

(a)

in outPS beads

Raman

(b)

PS bead

Laserspot

（2Oglycerol （2Oglycerol

(c)

cps

140012001000800600400200

0

700 750 800 850 900 950 1000

lowast

lowast


(d)

0

1000

800

600

400

200

800700600500400300200100

time (s)

lowast lowast

lowast

cps

at688

＝Ｇ-1

(e)

Figure 6 (a) TEM image of a AuNS targeted PS bead little particles can be observed on the curvature which are identified as AuNS in theinset magnifications (b) Sketch of the objective of the 120583-Raman over the microchannel (c) Optical microscope images of the microfluidicdevice A PS bead is caught just before entering on the sampling area (left) The elongated shape of the laser spot allows getting SERS signalsfrom the AuNS-PS bead for a prolonged period (right) (d) Raman spectrum recorded at 815 s where both the SERS-AuNS signal at 688 cmminus1and the PDMS peak at 710 cmminus1 are observed (e) Signals recorded at the fixed Raman shift of 688 cmminus1 during the flowing of the nanoparticlesin the microchannel

from the targeting of these objects can be detected with astandard 120583-Raman spectrometerThe two procedures presentdifferent advantages and disadvantages In particular theuse of a viscous liquid as carrier fluid allows reducing thebeadsrsquo speed similarly to what was obtained by introducinga downstream larger chamber with water or physiologicalbuffers but ensuring better conditions for the confinementThe combination of the two approaches can be exploited tofurther reduce the speed down to 10 times with respect to theinitial values Overall the most appropriate flow condition

should be considered depending on specific applicationsThepossibility of doing multiplexing analysis with SERS signalsexcited with only one laser line was already proved andshowed the potentiality of such a microfluidic device withrespect to a cytofluorimetric device

Data Availability

Data are available on request


Disclosure

Greshia Cappozzo present address is as follows StevanatoGroup Nuova Ompi Srl Via Molinella 17 35017 PiombinoDese (PD) Italy

Conflicts of Interest

The authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

The authors kindly acknowledge Daniele Filippi Dr PaoloSartori Dr Ladislav Derzsi and Giorgio Delfitto for usefuldiscussions and help in the experimental activities LucioLitti and Moreno Meneghetti would like to acknowledgethe University of Padova funding P-DiSC no 04BIRD2016-UNIPDand the Strategic Programof theUniversity of PadovaNAMECA

Supplementary Materials

Movie S1 polystyrene beads flowing in the flow-focusingmicrofluidic device Beadsrsquo path is well defined in the narrowchannel by the flow-focusing while it is less demarcatedin the chamber The movie is slowed down 15 times(Supplementary Materials)

References

[1] S Ding E You Z Tian and M Moskovits ldquoElectromagnetictheories of surface-enhanced Raman spectroscopyrdquo ChemicalSociety Reviews vol 46 no 13 pp 4042ndash4076 2017

[2] M Meneghetti A Scarsi L Litti et al ldquoPlasmonic nanostruc-tures for SERRS multiplexed identification of tumor-associatedantigensrdquo Small vol 8 no 24 pp 3733ndash3738 2012

[3] F Biscaglia S Rajendran P Conflitti et al ldquoEnhanced EGFRtargeting activity of plasmonic nanostructures with engineeredGE11 peptiderdquo Advanced Healthcare Materials vol 6 no 3Article ID 1700596 2017

[4] G MWhitesides ldquoThe origins and the future of microfluidicsrdquoNature vol 442 no 7101 pp 368ndash373 2006

[5] Y Xia and G M Whitesides ldquoSoft lithographyrdquo Annual Reviewof Materials Research vol 28 no 1 pp 153ndash184 1998

[6] E Piccin D Ferraro P Sartori E Chiarello M Piernoand G Mistura ldquoGeneration of water-in-oil and oil-in-watermicrodroplets in polyester-toner microfluidic devicesrdquo Sensorsand Actuators B Chemical vol 196 pp 525ndash531 2014

[7] S Silvestrini D Ferraro T Toth et al ldquoTailoring the wettingproperties of thiolene microfluidic materialsrdquo Lab on a Chipvol 12 no 20 pp 4041ndash4043 2012

[8] D Ferraro Y Lin B Teste et al ldquoContinuous chemical oper-ations and modifications on magnetic 120574-Fe

2O3nanoparticles

confined in nanoliter droplets for the assembly of fluorescentand magnetic SiO2120574-Fe2O3rdquo Chemical Communications vol51 no 95 pp 16904ndash16907 2015

[9] N Aboud D Ferraro M Taverna S Descroix C Smadja andN Thuy Tran ldquoDyneon THV a fluorinated thermoplastic as a

novel material for microchip capillary electrophoresisrdquoAnalystvol 141 no 20 pp 5776ndash5783 2016

[10] E K Sackmann A L Fulton and D J Beebe ldquoThe present andfuture role of microfluidics in biomedical researchrdquoNature vol507 no 7491 pp 181ndash189 2014

[11] M Serra D Ferraro I Pereiro J-L Viovy and S DescroixldquoThe power of solid supports in multiphase and droplet-basedmicrofluidics Towards clinical applicationsrdquoLab on aChip vol17 no 23 pp 3979ndash3999 2017

[12] D Ferraro J Champ B Teste et al ldquoMicrofluidic platformcombining droplets and magnetic tweezers Application toHER2 expression in cancer diagnosisrdquo Scientific Reports vol 6Article ID 25540 2016

[13] C D Ahrberg A Manz and B G Chung ldquoPolymerase chainreaction in microfluidic devicesrdquo Lab on a Chip vol 16 no 20pp 3866ndash3884 2016

[14] A K White M VanInsberghe O I Petriv et al ldquoHigh-throughput microfluidic single-cell RT-qPCRrdquo Proceedings ofthe National Acadamy of Sciences of the United States of Americavol 108 no 34 pp 13999ndash14004 2011

[15] I Hernandez-Neuta I Pereiro A Ahlford et al ldquoMicrofluidicmagnetic fluidized bed for DNA analysis in continuous flowmoderdquo Biosensors and Bioelectronics vol 102 pp 531ndash539 2018

[16] T D Mai D Ferraro N Aboud et al ldquoSingle-step immunoas-says and microfluidic droplet operation Towards a versatileapproach for detection of amyloid-120573 peptide-based biomarkersof Alzheimerrsquos diseaserdquo Sensors and Actuators B Chemical vol255 pp 2126ndash2135 2018

[17] G Wang C Das B Ledden Q Sun C Nguyen and SKumar ldquoEvaluation of disposable microfluidic chip design forautomated and fast Immunoassaysrdquo Biomicrofluidics vol 11 no1 Article ID 014115 2017

[18] B Teste N Jamond D Ferraro J-L Viovy and L MalaquinldquoSelective handling of droplets in a microfluidic device usingmagnetic railsrdquoMicrofluidics and Nanofluidics vol 19 no 1 pp141ndash153 2015

[19] LMou andX Jiang ldquoMaterials formicrofluidic immunoassaysa reviewrdquo Advanced Healthcare Materials vol 6 no 15 ArticleID 1601403 2017

[20] LMazutis J GilbertW L Ung D AWeitz A D Griffiths andJ A Heyman ldquoSingle-cell analysis and sorting using droplet-basedmicrofluidicsrdquoNature Protocols vol 8 no 5 pp 870ndash8912013

[21] H N Joensson and H Andersson Svahn ldquoDroplet microfluid-ics-A tool for single-cell analysisrdquo Angewandte Chemie Interna-tional Edition vol 51 no 49 pp 12176ndash12192 2012

[22] L W Yap H Chen Y Gao et al ldquoBifunctional plasmonic-magnetic particles for an enhanced microfluidic SERSimmunoassayrdquo Nanoscale vol 9 no 23 pp 7822ndash7829 2017

[23] C Wang F Madiyar C Yu and J Li ldquoDetection of extremelylow concentration waterborne pathogen using a multiplexingself-referencing SERSmicrofluidic biosensorrdquo Journal of Biolog-ical Engineering vol 11 article 9 2017

[24] Z Zhai F Zhang X Chen et al ldquoUptake of silver nanoparticlesby DHA-treated cancer cells examined by surface-enhancedRaman spectroscopy in amicrofluidic chiprdquo Lab on a Chip vol17 no 7 pp 1306ndash1313 2017

[25] K Kalantar-Zadeh K Khoshmanesh A A Kayani S Naha-vandi and A Mitchell ldquoDielectrophoretically tuneable opti-cal waveguides using nanoparticles in microfluidicsrdquo AppliedPhysics Letters vol 96 no 10 Article ID 101108 2010


[26] C Zhang K Khoshmanesh A Mitchell and K Kalantar-Zadeh ldquoDielectrophoresis for manipulation of micronanoparticles in microfluidic systemsrdquo Analytical and BioanalyticalChemistry vol 396 no 1 pp 401ndash420 2010

[27] A F Chrimes A A Kayani K Khoshmanesh et al ldquoDielec-trophoresis-Raman spectroscopy system for analysing sus-pended nanoparticlesrdquo Lab on a Chip vol 11 no 5 pp 921ndash9282011

[28] A F Chrimes K Khoshmanesh S-Y Tang et al ldquoIn situ SERSprobing of nano-silver coated individual yeast cellsrdquo Biosensorsand Bioelectronics vol 49 pp 536ndash541 2013

[29] A F Chrimes K Khoshmanesh P R Stoddart et al ldquoActivecontrol of silver nanoparticles spacing using dielectrophoresisfor surface-enhanced Raman scatteringrdquo Analytical Chemistryvol 84 no 9 pp 4029ndash4035 2012

[30] J Oakey RW Applegate E Arellano D D Carlo SW Gravesand M Toner ldquoParticle focusing in staged inertial microfluidicdevices for flow cytometryrdquo Analytical Chemistry vol 82 no 9pp 3862ndash3867 2010

[31] J-C Baret O J Miller V Taly et al ldquoFluorescence-activateddroplet sorting (FADS) Efficient microfluidic cell sorting basedon enzymatic activityrdquo Lab on a Chip vol 9 no 13 pp 1850ndash1858 2009

[32] A Pallaoro M R Hoonejani G B Braun C D Meinhartand M Moskovits ldquoRapid identification by surface-enhancedraman spectroscopy of cancer cells at low concentrationsflowing in a microfluidic channelrdquo ACS Nano vol 9 no 4 pp4328ndash4336 2015

[33] E Locatelli M Pierno F Baldovin E Orlandini Y Tan andS Pagliara ldquoSingle-File Escape of Colloidal Particles fromMicrofluidic Channelsrdquo Physical Review Letters vol 117 no 3Article ID 038001 2016

[34] A Karimi S Yazdi and A M Ardekani ldquoHydrodynamicmechanisms of cell and particle trapping in microfluidicsrdquoBiomicrofluidics vol 7 no 2 Article ID 021501 2013

[35] X Xuan J Zhu and C Church ldquoParticle focusing in microflu-idic devicesrdquoMicrofluidics and Nanofluidics vol 9 no 1 pp 1ndash16 2010

[36] P Tabeling Introduction to Microfluidics Oxford UniversityPress 2005 httpsbooksgoogleitbooksaboutIntroductionto Microfluidicshtmlid=h4ZguvxYW0kCampampredir esc=y

[37] D Ferraro M Serra I Ferrante J-L Viovy and S DescroixldquoMicrofluidic valve with zero dead volume and negligible back-flow for droplets handlingrdquo Sensors and Actuators B Chemicalvol 258 pp 1051ndash1059 2018

[38] W Tan and S Takeuchi ldquoA trap-and-release integratedmicrofluidic system for dynamic microarray applicationsrdquo Pro-ceedings of the National Acadamy of Sciences of the United Statesof America vol 104 no 4 pp 1146ndash1151 2007

[39] S Kobel A Valero J Latt P Renaud andM Lutolf ldquoOptimiza-tion of microfluidic single cell trapping for long-term on-chipculturerdquo Lab on a Chip vol 10 no 7 pp 857ndash863 2010

[40] E Chiarello A Gupta G Mistura M Sbragaglia and MPierno ldquoDroplet breakup driven by shear thinning solutions inamicrofluidic T-junctionrdquo Physical Review Fluids vol 2 ArticleID 123602 2017

[41] V Amendola L Litti and M Meneghetti ldquoLDI-MS assistedby chemical-free gold nanoparticles Enhanced sensitivity andreduced background in the low-mass regionrdquo Analytical Chem-istry vol 85 no 24 pp 11747ndash11754 2013

[42] L Litti N Rivato G Fracasso et al ldquoA SERRSMRImultimodalcontrast agent based on naked Au nanoparticles functionalizedwith a Gd(iii) loaded PEG polymer for tumor imaging andlocalized hyperthermiardquoNanoscale vol 10 no 3 pp 1272ndash12782018

[43] M R Hoonejani A Pallaoro G B Braun M Moskovitsand C D Meinhart ldquoQuantitative multiplexed simulated-cellidentification by SERS in microfluidic devicesrdquo Nanoscale vol7 no 40 pp 16834ndash16840 2015

[44] F Lamberti L Litti M De Bastiani et al ldquoHigh-QualityLigands-Free Mixed-Halide Perovskite Nanocrystals Inks forOptoelectronic Applicationsrdquo Advanced Energy Materials vol7 no 8 2017

[45] F Bertorelle M Pinto R Zappon et al ldquoSafe core-satellitemagneto-plasmonic nanostructures for efficient targeting andphotothermal treatment of tumor cellsrdquo Nanoscale vol 10 no3 pp 976ndash984 2018

[46] V Amendola S Scaramuzza L Litti et al ldquoMagneto-plasmonicAu-Fe alloy nanoparticles designed formultimodal SERS-MRI-CT imagingrdquo Small vol 10 no 12 pp 2476ndash2486 2014

[47] G Sciutto L Litti C Lofrumento et al ldquoAlternative SERRSprobes for the immunochemical localization of ovalbumin inpaintings An advanced mapping detection approachrdquo Analystvol 138 no 16 pp 4532ndash4541 2013

Hindawiwwwhindawicom Volume 2018

Active and Passive Electronic Components


Shock and Vibration


High Energy PhysicsAdvances in

Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom

The Scientific World Journal

Volume 2018

Acoustics and VibrationAdvances in



Advances in Condensed Matter Physics

OpticsInternational Journal of



AstronomyAdvances in

Antennas andPropagation

International Journal of




Geophysics

Advances inOpticalTechnologies

Hindawiwwwhindawicom

Volume 2018

Applied Bionics and BiomechanicsHindawiwwwhindawicom Volume 2018

Advances inOptoElectronics


Volume 2018


Mathematical PhysicsAdvances in


ChemistryAdvances in


Journal of

Chemistry


Advances inPhysical Chemistry


RotatingMachinery



Journal ofEngineeringVolume 2018

Submit your manuscripts atwwwhindawicom


Examples of microfluidic devices coupled with Ramanspectroscopy have been successfully used for sample precon-centration or storing cells during the spectroscopic analysis[22ndash24]

For example dielectrophoresis (DIE) is largely used formanipulating particles in microfluidic devices [25 26] andmore recently this approach has been recently combinedwithSERS analysis [27] for detection of specific biomarkers [28]and cells [29] However although DIE offers the advantageof precisely positioning cells or particles in the channel itrequires devices with integrated electrodes showing specificgeometry increasing the complexity and the cost of themicrofabrication processes and preventing the productionof disposable device [30] A passive approach that requiressimpler fabrication protocols is represented by the use of aflow-focusing geometry to align cells (or particles) in a singleline and force them to pass under the Raman laser spotssimilarly to what is shown with UV-visible spectroscopy inflow [31]However only onework has been presented so far topush forward the flow-focusing geometry with SERS analysis[32] In fact while this device allowed acquiring continuouslythe signal from flowing cells it presents twomain drawbacksfirst the usability of this procedure is severely limited bythe acquisition time of SERS spectrometers (given by theCCD exposure and shutter) In addition even with fastspectrometers (100-150ms of acquisition time) it is necessaryto design themicrofluidic devices to reduce the velocity of thecarrying flow Since this achievement is commonly obtainedby reenlarging the carrying flow after single file confinement[33] it has been demonstrated that beads flowing in amicrofluidic device can be laterally displaced according totheir size and stiffness [34 35] This may introduce someuncontrolled deviation from the laser spot an issue whendealing with clinical analysis

Given that the fluidic manipulation of cells (or beads) atslow speed in amicrofluidic device is highly demanding sinceit requiresworking at very lowpressure (typically hundreds ofPa) causing nonstable flows a different approach can exploitthe viscosity 120583 of the carrier fluid solution

Within this approach it is possible to affect the hydraulicresistance 119877ℎ of the microfluidic device and so the final flowrate As a matter of fact in a microchannel once fixed thepressure drop Δ119875 at the channel ends and the flow rate 120593 canbe calculated as [36]

120593 =Δ119875

119877ℎ(1)

119877ℎ for a microfluidic channel of length 119871 and radius 119903 can beapproximated as [37]

119877ℎ =8120583119871

1205871199034(2)

Therefore with fixed geometry an increase of 120583 leads to aproportional increase of119877ℎ and thus a reduction of the globalflow rate in the device (see (1))

We believe that both strategies for reducing the cells speedare interesting and both show advantages and disadvantages(i) changing the geometry is very simple in the fabrication

step however it leads to a nonperfect straight trajectory ofthe cells (ii) acting on the viscosity of the liquid could be notalways possible however it permits working with a narrowedconfined flow and thus a deterministic path of the cells Inthis paper we will compare the two strategies demonstratinghow even the latter one can be exploited to perform SERSin a microfluidic device We used for proof-of-conceptexperiments biotinylated polystyrene beads (PS beads) ascell model [38 39] SERS nanostructures produced withgold nanoparticles functionalized with an optimized SERSreporter (NPCy(SH)2) were also functionalized with strep-tavidin The biotin presented on the PS beads and the strep-tavidin on the gold nanostructures represent the antigen-antibody pair used for the identification of cancer cells

First we introduce the methods for the fabrication ofthe microfluidic devices the production of the SERS activenanostructures and the experimental setup Then we reportthe numerical and experimental results for the fluidic charac-terization and we demonstrate the possibility of performingSERS analysis of beads flowing in the microfluidic deviceFinally we conclude comparing the two fluidic strategies andthe results

2 Materials and Methods

21 Microfluidic Device Fabrication and LiquidHandling Themicrofluidic device was produced by conventional replicamolding [5 40] with polydimethylsiloxane (or PDMS Syl-gard 184 by Dow Corning) of a mold prepared by pho-tolithography with SU-8 (by MicroChem) In detail afterhaving produced the mold the nonpolymerized PDMS waspoured on it and placed in an oven at 70∘C for 1h Afterthat the PDMS part was peeled off punched for inlets andbonded to a glass slide by Oxygen Plasma treatment FinallyPTFE tubing having 03mm (06mm) inner (outer) diameter(by Sigma Aldrich) were fixed by epoxy glue as shown inFigure 1(a) The final design described in Figure 1(b) showsa flow-focusing junction aimed at creating a focused flow inthe center channel The device presents two inletsrsquo channelone for the carrier liquid used to generate the confined flowand the other for the liquids containing the functionalizedbeads both have the same cross section 100120583m times20120583m oflateral and thickness dimensions respectivelyThen after thejunction the narrow channel opened in a larger chamber(w=700120583m) aimed at reducing the flow rate and thus thespeed of the beads However as introduced before the con-fined stream flow in the chamber is larger than in the channel(see Figure 1(b)) and thus the beads position becomes lessdetermined (see also Movie S1 in Supplementary Materials)

In order to optimize the focusing condition and charac-terize the speed of the beads different glycerolwater mixeswere used as carrier fluid varying the viscosity between 1 and120 cP then during the flow characterization experimentsthe inner fluid was composed of non-coated polystyrene(PS) beads (size 80-129120583m by Spherotech) dispersed in aPercollwater solution (by GE Healthcare) at 42 ww Thelatter was used to match the density of the PS beads in orderto reduce the sedimentation


PDMS

glass slide

1 23

PTFE tubing

(a)

100m

Flow Floww

beads

carrier fluid

inner fluid

D

０1

０2０3

＄

(b)













(a)

120

100

80

60

40

20

0

D [

m]

00 02 04 06 08 10 12 14

０2０1

(b)


distance (mm)

w

local speed(ms)

dist

ance

(mm

)

r=100m

200m

(a)

15

14

13

12

11

10

09

08

07

06

05

Nor

mal

ized

spee

d (v

Ｐ )

0 100 200 300 400 500 600 700 800 900 1000

w (m)

(b)






10

9

8

7

6

5

4

3

2

1

0

Spee

d (m

ms

)

0 50 100 150 200 250

viscosity (cP)

visc

osity

(cP)

channelchamber

glycerol conc ()

250

200

150

100

50

0

0 20 40 60 80 100













AuNPby LASiS

AuNS

０Ｓ(３（)2（2（2３（

（2（2３（

N

NN

N

N

NN

NSi

(a)05

04

03

02

01

00

＜Ｍ＜

=2ＧＧ

300 400 500 600 700 800 900 1000 1100

Wavelength (nm)

(b)

cps

1800

1600

1400

1200

1000

800

600

400

200

1800 20001600140012001000800600400200

0


688 ＝Ｇ-1

(c)





4 Conclusions



PS bead

AuNS

(a)

in outPS beads

Raman

(b)

PS bead

Laserspot


(c)

cps

140012001000800600400200

0

700 750 800 850 900 950 1000

lowast

lowast


(d)

0

1000

800

600

400

200

800700600500400300200100

time (s)

lowast lowast

lowast

cps

at688

＝Ｇ-1

(e)




Data Availability



Disclosure




Acknowledgments




References









2O3nanoparticles














































Shock and Vibration





Volume 2018














Geophysics



Volume 2018




Volume 2018






Journal of

Chemistry




RotatingMachinery






PDMS

glass slide

1 23

PTFE tubing

(a)

100m

Flow Floww

beads

carrier fluid

inner fluid

D

０1

０2０3

＄

(b)













(a)

120

100

80

60

40

20

0

D [

m]

00 02 04 06 08 10 12 14

０2０1

(b)


distance (mm)

w

local speed(ms)

dist

ance

(mm

)

r=100m

200m

(a)

15

14

13

12

11

10

09

08

07

06

05

Nor

mal

ized

spee

d (v

Ｐ )

0 100 200 300 400 500 600 700 800 900 1000

w (m)

(b)






10

9

8

7

6

5

4

3

2

1

0

Spee

d (m

ms

)

0 50 100 150 200 250

viscosity (cP)

visc

osity

(cP)

channelchamber

glycerol conc ()

250

200

150

100

50

0

0 20 40 60 80 100













AuNPby LASiS

AuNS

０Ｓ(３（)2（2（2３（

（2（2３（

N

NN

N

N

NN

NSi

(a)05

04

03

02

01

00

＜Ｍ＜

=2ＧＧ

300 400 500 600 700 800 900 1000 1100

Wavelength (nm)

(b)

cps

1800

1600

1400

1200

1000

800

600

400

200

1800 20001600140012001000800600400200

0


688 ＝Ｇ-1

(c)





4 Conclusions



PS bead

AuNS

(a)

in outPS beads

Raman

(b)

PS bead

Laserspot


(c)

cps

140012001000800600400200

0

700 750 800 850 900 950 1000

lowast

lowast


(d)

0

1000

800

600

400

200

800700600500400300200100

time (s)

lowast lowast

lowast

cps

at688

＝Ｇ-1

(e)




Data Availability



Disclosure




Acknowledgments




References









2O3nanoparticles














































Shock and Vibration





Volume 2018














Geophysics



Volume 2018




Volume 2018






Journal of

Chemistry




RotatingMachinery






(a)

120

100

80

60

40

20

0

D [

m]

00 02 04 06 08 10 12 14

０2０1

(b)


distance (mm)

w

local speed(ms)

dist

ance

(mm

)

r=100m

200m

(a)

15

14

13

12

11

10

09

08

07

06

05

Nor

mal

ized

spee

d (v

Ｐ )

0 100 200 300 400 500 600 700 800 900 1000

w (m)

(b)






10

9

8

7

6

5

4

3

2

1

0

Spee

d (m

ms

)

0 50 100 150 200 250

viscosity (cP)

visc

osity

(cP)

channelchamber

glycerol conc ()

250

200

150

100

50

0

0 20 40 60 80 100













AuNPby LASiS

AuNS

０Ｓ(３（)2（2（2３（

（2（2３（

N

NN

N

N

NN

NSi

(a)05

04

03

02

01

00

＜Ｍ＜

=2ＧＧ

300 400 500 600 700 800 900 1000 1100

Wavelength (nm)

(b)

cps

1800

1600

1400

1200

1000

800

600

400

200

1800 20001600140012001000800600400200

0


688 ＝Ｇ-1

(c)





4 Conclusions



PS bead

AuNS

(a)

in outPS beads

Raman

(b)

PS bead

Laserspot


(c)

cps

140012001000800600400200

0

700 750 800 850 900 950 1000

lowast

lowast


(d)

0

1000

800

600

400

200

800700600500400300200100

time (s)

lowast lowast

lowast

cps

at688

＝Ｇ-1

(e)




Data Availability



Disclosure




Acknowledgments




References









2O3nanoparticles














































Shock and Vibration





Volume 2018














Geophysics



Volume 2018




Volume 2018






Journal of

Chemistry




RotatingMachinery






10

9

8

7

6

5

4

3

2

1

0

Spee

d (m

ms

)

0 50 100 150 200 250

viscosity (cP)

visc

osity

(cP)

channelchamber

glycerol conc ()

250

200

150

100

50

0

0 20 40 60 80 100













AuNPby LASiS

AuNS

０Ｓ(３（)2（2（2３（

（2（2３（

N

NN

N

N

NN

NSi

(a)05

04

03

02

01

00

＜Ｍ＜

=2ＧＧ

300 400 500 600 700 800 900 1000 1100

Wavelength (nm)

(b)

cps

1800

1600

1400

1200

1000

800

600

400

200

1800 20001600140012001000800600400200

0


688 ＝Ｇ-1

(c)





4 Conclusions



PS bead

AuNS

(a)

in outPS beads

Raman

(b)

PS bead

Laserspot


(c)

cps

140012001000800600400200

0

700 750 800 850 900 950 1000

lowast

lowast


(d)

0

1000

800

600

400

200

800700600500400300200100

time (s)

lowast lowast

lowast

cps

at688

＝Ｇ-1

(e)




Data Availability



Disclosure




Acknowledgments




References









2O3nanoparticles














































Shock and Vibration





Volume 2018














Geophysics



Volume 2018




Volume 2018






Journal of

Chemistry




RotatingMachinery







AuNPby LASiS

AuNS

０Ｓ(３（)2（2（2３（

（2（2３（

N

NN

N

N

NN

NSi

(a)05

04

03

02

01

00

＜Ｍ＜

=2ＧＧ

300 400 500 600 700 800 900 1000 1100

Wavelength (nm)

(b)

cps

1800

1600

1400

1200

1000

800

600

400

200

1800 20001600140012001000800600400200

0


688 ＝Ｇ-1

(c)





4 Conclusions



PS bead

AuNS

(a)

in outPS beads

Raman

(b)

PS bead

Laserspot


(c)

cps

140012001000800600400200

0

700 750 800 850 900 950 1000

lowast

lowast


(d)

0

1000

800

600

400

200

800700600500400300200100

time (s)

lowast lowast

lowast

cps

at688

＝Ｇ-1

(e)




Data Availability



Disclosure




Acknowledgments




References









2O3nanoparticles














































Shock and Vibration





Volume 2018














Geophysics



Volume 2018




Volume 2018






Journal of

Chemistry




RotatingMachinery






PS bead

AuNS

(a)

in outPS beads

Raman

(b)

PS bead

Laserspot


(c)

cps

140012001000800600400200

0

700 750 800 850 900 950 1000

lowast

lowast


(d)

0

1000

800

600

400

200

800700600500400300200100

time (s)

lowast lowast

lowast

cps

at688

＝Ｇ-1

(e)




Data Availability



Disclosure




Acknowledgments




References









2O3nanoparticles














































Shock and Vibration





Volume 2018














Geophysics



Volume 2018




Volume 2018






Journal of

Chemistry




RotatingMachinery






Disclosure




Acknowledgments




References









2O3nanoparticles














































Shock and Vibration





Volume 2018














Geophysics



Volume 2018




Volume 2018






Journal of

Chemistry




RotatingMachinery































Shock and Vibration





Volume 2018














Geophysics



Volume 2018




Volume 2018






Journal of

Chemistry




RotatingMachinery








Shock and Vibration





Volume 2018














Geophysics



Volume 2018




Volume 2018






Journal of

Chemistry




RotatingMachinery





Documents

Single File Flow of Biomimetic Beads for Continuous SERS ...downloads.hindawi.com/journals/acmp/2018/2849175.pdf · AdvancesinCondensedMatterPhysics 10 9 8 7 6 5 4 3 2 1 0 Speed (mm/s)