The Effect of the Orientation Towards Analyte Flow on Electrochemical Sensor Performance and Current Fluctuations
Abstract
Analyte flow influences the performance of every gas sensor; thus, most of these sensors usually contain a diffusion barrier (layer, cover, inlet) that can prevent the negative impact of a sudden change of direction and/or the rate of analyte flow, as well as various unwanted impacts from the surrounding environment. However, several measurement techniques use the modulation of the flow rate to enhance sensor properties or to extract more information about the chemical processes that occur on a sensitive layer or a working electrode. The paper deals with the experimental study on how the analyte flow rate and the orientation of the electrochemical sensor towards the analyte flow direction influence sensor performance and current fluctuations. Experiments were carried out on a semi-planar, three-electrode topology that enabled a direct exposure of the working (sensing) electrode to the analyte without any artificial diffusion barrier. The sensor was tested within the flow rate range of 0.1–1 L/min and the orientation of the sensor towards the analyte flow direction was gradually set to the four angles 0°, 45°, 90° and 270° in the middle of the test chamber, while the sensor was also investigated in the standard position at the bottom of the chamber.
Full text
Sensors2020,20,1038;doi:10.3390/s20041038www.mdpi.com/journal/sensors Article TheEffectoftheOrientationTowardsAnalyteFlow onElectrochemicalSensorPerformanceandCurrent Fluctuations PetrSedlák1,*andPetrKuberský2 1FacultyofElectricalEngineeringandCommunications,BrnoUniversityofTechnology,Technická10, Brno61600,CzechRepublic 2FacultyofElectricalEngineering,RegionalInnovationCentreforElectricEngineering, UniversityofWestBohemia,Univerzitni8,Plzen30100,CzechRepublic;[email protected]u.cz *Correspondence:[email protected];Tel.:+420‐541‐146‐021 Received:21January2020;Accepted:13February2020;Published:14February2020 Abstract:Analyteflowinfluencestheperformanceofeverygassensor;thus,mostofthesesensors usuallycontainadiffusionbarrier(layer,cover,inlet)thatcanpreventthenegativeimpactofa suddenchangeofdirectionand/ortherateofanalyteflow,aswellasvariousunwantedimpacts fromthesurroundingenvironment.However,severalmeasurementtechniquesusethemodulation oftheflowratetoenhancesensorpropertiesortoextractmoreinformationaboutthechemical processesthatoccuronasensitivelayeroraworkingelectrode.Thepaperdealswiththe experimentalstudyonhowtheanalyteflowrateandtheorientationoftheelectrochemicalsensor towardstheanalyteflowdirectioninfluencesensorperformanceandcurrentfluctuations. Experimentswerecarriedoutonasemi‐planar,three‐electrodetopologythatenabledadirect exposureoftheworking(sensing)electrodetotheanalytewithoutanyartificialdiffusionbarrier. Thesensorwastestedwithintheflowraterangeof0.1–1L/minandtheorientationofthesensor towardstheanalyteflowdirectionwasgraduallysettothefourangles0°,45°,90°and270°inthe middleofthetestchamber,whilethesensorwasalsoinvestigatedinthestandardpositionatthe bottomofthechamber. Keywords:analyteflow;currentfluctuations;amperometricsensor;signal‐to‐noiseratio 1.Introduction Chemicalgassensorsrepresentlow‐cost,easy‐to‐usedeviceswithahighminiaturization potentialtodetectchemicalsubstancesdispersedinanenvironmentforawidespectrumof applications[1,2].Thesensormechanismsandtheirpropertiesaretightlyboundwithelectriccharge transportat/inactiveelectrochemicalinterfaces[3],andaredeliberatelyinfluencedbychanging workingconditionsorunintentionallybytheoutsideenvironment.Itshouldbenotedthatseveral measurementtechniquesusemodulationoftheflowrate[4–7],thetemperature[8–12]oreventhe light[13–15]toextractmoreinformationaboutthechemicalprocessesortoenhancesensor properties,sincetheelectrochemicalsensor(suchasaconductometricoramperometricone)could exhibitcross‐interference,i.e.,non‐selectivity,forchemicallysimilarmolecules[1,16]. Theanalyteflowrateinfluencestheperformanceofeverygassensor;thus,mostofthesesensors usuallycontainadiffusionbarrier(layer,coverorinlet)thatmaypreventthenegativeimpactofa suddenchangeofdirectionand/ortherateofanalyteflow,aswellasvariousunwantedimpactsfrom thesurroundingenvironment[17].Severalauthorsdescribedtheeffectoftheflowrateonthesignal andsensorpropertiesofgassensors[18–21].Ourpreviouspaper[21]demonstratedonafully‐printed
Sensors2020,20,10382of17 amperometricgassensorthatthedirectcurrent(DC)throughthesensorexhibitschangeswithinthe rangeofoneorderforanon‐zeroflowrate,whilethespectraldensityofcurrentfluctuations significantlychangesitsshapeandthelevelofcurrentfluctuationsinthepowerspectrumforthe ordersastheflowrateincreasesataconstantconcentrationofthedetectedgas.Furthermore,the signal‐to‐noiseratioofresponsewasalmostinvarianttogasconcentrationanddecreasedwithflow rateincrease.Theseresultsindicatetheimportanceofcurrentfluctuationmeasurements,makingit possibletoextractnewandvaluableinformation(parameters)resultinginafluctuation‐enhanced detectiontechnique. Thispaperdealswiththeexperimentalstudyofhowtheanalyteflowrateandtheorientation ofthesensor,i.e.,thesurfaceareaoftheworkingelectrode(WE)towardstheanalyteflowdirection, impactoverallsensorperformanceandcurrentfluctuations.Thefirstsetofexperimentswasaimed atacomparisonofselectedsensorparameters(sensitivity,response/recoverytime,limitofdetection andrepeatability)dependentonboththesizeoftheWEsurfaceareaandtheanalyteflowrate.Within thesecondsetofexperiments,attentionwaspaidtotheeffectofthemutualanglebetweentheanalyte flowdirectionandtheWEsurfaceareaonsensorparameters,wheretheorientationofthesensor towardstheanalyteflowwasgraduallysettothefourangles0°,45°,90°and270°inthemiddleof thetestchamber.Thethirdsetofexperimentswasaimedatthestudyofcurrentfluctuations dependingonsensorrotationfortheaforementionedanglesbetweentheanalyteflowdirectionand thesurfaceareaoftheworkingelectrode.Twodependenceswerestudiedseparatelyforeachangle: (i)theeffectofanalyteconcentrationataparticularflowrate,(ii)theeffectofanalyteflowrateata particularconcentration. 2.MaterialsandMethods 2.1.SensorPreparation Allexperimentswerecarriedoutonawell‐established,three−electrodesensorplatform(see Figure1)basedonaceramicsubstratewithaplatinumcounter(CE)andaplatinumpseudoreference electrode(RE).Asolidpolymerelectrolyte(SPE)contained1‐ethyl‐3‐methylimidazolium bis(trifluoromethylsulfonyl)imide[C 2 mim][NTf 2 ]ionicliquid,poly(vinylidenefluoride)(PVDF)and 1‐methyl‐2‐pyrrolidone(NMP).Moreinformationaboutthesensorplatformandadetailed descriptionofSPEpreparationcanbefoundin[22–24]and[25,26],respectively.Acarbonworking electrodewaspreparedbyspraycoatingamixtureconsistingof300mgglassycarbonsphericalpowder (2−12μm,Sigma−Aldrich,Germany)and1mLofethanol.Threedifferentworkingelectrodeareas werepreparedwithregardtotheoveralldimensionofthebasicceramicsubstrate.Theratioof geometricareasbetweenWEandCEwereapproximately1/5,1/2and1fortheWEsurfacearea2.9, 8.5and22.4mm 2 ,respectively. Figure1.Sensortopologywiththreedifferentworkingelectrodeareas. 2.2.ExperimentalSetup TheapparatusforsensortestingconsistedofPC‐controlledmassflowcontrollers (SmartTrak100,SierraInstruments,Monterey,USA‐California),gascylinderswithsyntheticairand
Sensors2020,20,10383of17 areferencecalibrationmixture(100ppmNO 2 balancedinnitrogen),atestchamberandreadout electronics.Forfurtherdescriptionoftheapparatus,theauthorsrefertoapreviousstudy[21].For thefirstsetofexperiments,whenthetestedsensorswereplacedontothetestchamberbottom,gold springprobeswereusedtocontactallthreeelectrodes.Forthesecondandthirdsetofexperiments, thetestedsensorwascontactedviasilverwiresthatallowedtherotationofthesensor.A programmableanalogfront‐end(AFE)potentiostatLMP91000(TexasInstrument,Dallas,USA)was usedasreadoutelectronicsforallthemeasurementsexceptthemeasurementofcurrentfluctuation. BiasvoltagebetweenREandWEwassetto0.5Vforalltheexperimentsandthesensorresponse currentwasrecordedeveryothersecond. Themeasurementsetupforevaluationofcurrentfluctuationwerebasedonourownbattery‐fed device,whichimplementsthepotentiostatcircuitinagrounded‐WEconfiguration[27]andalow‐noise transimpedanceamplifierwithseparatedalternatingcurrent(AC)andDCoutputs.Thesensorunder testisconnectedtothepotentiostatcircuit.TheACvoltageoutputwasledtoanamplifierwithhighly selectivefiltersAM22(3SSedlak,s.r.o.,Brno,CzechRepublic),andwasobtainedbya12‐bitAD convertorHS5(TiePieengineering,Sneek,Netherlands)aswellasDCvoltageoutput.Toeliminatethe externaldisturbancefromtheenvironment,thetestchamberwiththesensorandourdeviceislocated inafaradaycageandalldevicesofthemeasurementchainwerepoweredfrombatteries. 3.ResultsandDiscussion Toillustrateairflowbehaviorinthetestchamberatvariousflowrates,simplifiedpreliminary simulationswerecarriedoutwithouttheconsiderationofasensor,contactneedles,etc.Figure2a showscomputationalfluiddynamics(CFD)simulations[28]ofairflowinthegaschamberattheflow ratesof0.1L/min,0.5L/minand1.0L/min.Thecomputationwascarriedoutinthesoftware environmentANSYS17.0forthechamberwithoutanysensororcontactneedles.TheCFD simulationsshowthatvorticesareformedforallrates,thus,theflowinthechamberisturbulent[29].At 0.1L/min,theflowalongthelongitudinalmid‐planeexhibitminimalinfluencefromvortices,while forhigherflowratesthedegreeofturbulencegrows.Twosignificantvorticesareformed symmetricallytothelongitudinalplaneofthechamber,andchangetheirpositionsfromtheleftto therightastheflowrateincreases.Oneofthesetwovorticesissupposedtosignificantlycontribute tothesignalresponseofasensor,sinceitdirectlyinfluencestheflowaroundthesensor.Figure2b illustratesthepositionofthesensorduringthemeasurementsintheairaswellasstreamlinesatthe flowrateof1L/min.
Sensors2020,20,10384of17 Figure2.(a)Computationalfluiddynamics(CFD)simulationsofairflowinthetestchamberatflowrates 0.1L/min,0.5L/minand1.0L/min,(b)sensorpositionandorientationtowardstheflowatangle0° and90°inthegaschamber. Theorientationofthesensortowardstheanalyteflowwassetgraduallytothefouranglesof0°, 45°,90°and270°,whilethesensorwasalsoinvestigatedinthestandardpositionatthebottomofthe gaschamber[21].Accordingtosimulations,theanalyteflowfrominletseemstoimpactdirectlyWE electrodeofthesensoratangles0°,45°and90°.Detailedsimulationsofthefluidicconditionsinthe vicinityofthesensorwerenotcarriedoutduetotheirhighcomplexity.Figure3showsthelocation androtationofthesensorforthe0°angle. Figure3.Sensorpositioninthegaschamberatangle0°. 3.1.EffectoftheWorkingElectrode(WE)SurafceAreaandAnalyteFlowRateonSensorParameters Withinthefirstsetofexperiments,wetestedsensorswiththreedifferentworkingelectrode areas(2.9,8.5and22.4mm 2 )andcomparedtheirbasicsensorparameters(sensitivity, response/recoverytime,limitofdetection,repeatability)inordertodeterminetheinfluenceofa differentworkingelectrodeareaonsensorparameters.Eachsensorwasexposedtothesametest profilethatconsistedofastepwiseincreaseinnitrogendioxideconcentrationfrom0to3ppm(1step equaled1ppmNO 2 )withsubsequentthreeconsecutiveexposurestothesameconcentrationof3ppm NO 2 .Weappliedthesametestprofilewithfourdifferenttotalflowratesofanalyte(0.1,0.5,0.8and 1L/min)inordertoobservetheimpactoftheflowratelevelonsensorparametersforeachsizeof theworkingelectrodearea.Relativehumidityandtemperaturewereconstantwithinallexperiments (298Kand40%RHunlessotherwisestated)andeachofthesensorstestedwasplacedonthesame positioninthetestchambertobeexposedtothemostidenticalconditions.Figure4showstypical sensorresponsestothetestcycle.Sensitivitywasdeterminedastheslopeofacalibrationcurve.For theconstructionofthecalibrationcurve,anaveragevalueofsensorcurrentwasdeterminedforeach concentration.Averagevalues(illustratedbyred,greenandbluesquaresinFigure4)werecalculated fromthelastminuteofeachconcentrationlevel(highlightedbythegrayzone),wheresensor responseswerestable.Response/recoverytime(T 90 /T 10 )wascalculatedasthetimeperiodnecessary toachieve90%or10%ofthesteadystatecurrentuponastepincrease/decreaseinNO 2 concentration. Furthersensorparameters,thelimitofdetection(LOD)andrepeatability,willbediscussedinthe followingtext.
Sensors2020,20,10385of17 Figure4.Sensorresponsetothetestcycleforsampleswithdifferentworkingelectrode(WE)surface areas(conditions:298K,40%RH(relativehumidity),analyteflowrate0.5L/min). Figure5showssensitivitydependenceontheWEsurfaceareafordifferentanalyteflowrates (Figure5a)andonanalyteflowratesfordifferentWEsurfaceareas(Figure5b).Sensorsensitivity exhibitedalinearlygrowingtrendwiththeincreaseofboththeWEsurfaceareaandtheanalyteflow rate.Theincreaseofsensitivityperunitsurfaceareawas19,25.1,30.5and31.4nA/ppmforanalyte flowratesof0.1,0.5,0.8and1L/min,respectively.Thehighertheflowrate,thehigherthechangeof sensitivityperunitsurfacearea.Theincreaseof100mL/minofanalyteflowratecausedasensitivity increaseof3.9,9.8and31.5nA/ppmfortheWEsurfaceareaof2.9,8.5and22.4mm 2 ,respectively.The largertheWEsurfacearea,thehigherchangeofsensitivityperunitanalyteflowrate. Figure5.(a)SensitivitydependenceontheWEsurfacearea;(b)sensitivitydependenceontheanalyte flowrate.TheerrorbarsinFigure5representthestandarderroroftheslopeofthecalibrationcurve (sensitivity)thatwasconstructedbylinearregression. Figures6and7illustratetheresponse/recoverytimedependencesontheWEsurfaceareafor differentanalyteflowrates(Figures6aand7a)andontheanalyteflowratesfordifferentWEsurface areas(Figures6band7b).ItcanbereasonablysaidthattheinfluenceofWEsurfaceareaonboththe response/recoverytimewasnegligible.Differencesfromallmeasurementsfellwithintherangeof±2s withoutanycleartrendwithrespecttotheWEsurfacearea.Thisfactisnotsurprisingbecausethe areaoftheworkingelectrode,whichiscloselyrelatedwiththeareaoftheelectrochemicalactive interfacebetweentheSPElayerandtheworkingelectrode,usuallyinfluencethesensorsensitivity. Response/recoverytimeofthesensorisusuallyinfluencedbythetransportoftheanalytetothe electrochemicallyactive(electrode‐electrolyte)interface.Withinourexperiments,sensorswerenot intentionallyprotectedanydiffusionbarrierthatusuallycontrolsthetransportoftheanalytethrough thesensorhousingandsotheonlythicknessoftheworkingelectrode,throughwhichtheanalyte penetratedtotheinterfacewiththeSPE,couldsignificantlyinfluenceresponse/recoverytime.
Sensors2020,20,10386of17 AlthoughwedonothaveexactvaluesofWEthicknessesforparticularsensorswithdifferentWE areaswesupposethatparticularworkingelectrodeswerenearlythesamethicknesswhichresulted inthenegligibleeffectoftheWEsurfaceareaontheresponse/recoverytime.Figures6band7bshow responseandrecoverytimedependenceontheanalyteflowrate.Whileaslightincreaseinboth response/recoverytimewasobservedwhentheanalyteflowratewasdecreasedfrom1L/minto0.5 L/min,asubstantialincreaseinbothtimeswasobservedwhentheanalyteflowratewassetto0.1 L/min.However,thisfactcouldhavepartiallycontributedtothetimeperiodwhichisnecessaryto exchangetheinternaldeadvolumeofthetestchamber(90mL). TheLODcalculatedastheratioofthetriplestandarddeviationofthebackgroundcurrentnoise (atzeroconcentration)andsensitivityisarathertheoreticalvaluethatcanusuallybeverifiedwith difficulty.First,anaccurateandreproduciblepreparationofsuchlowconcentrationlevels(ideally equalingthecalculateddetectionlimits)isveryproblematic.Inourcase,thetestingapparatus allowedaminimumstepincreaseequaling100ppbNO 2, whichresultedinanaverageincreaseof83nA ofthereductioncurrentI CE withintherangeof0−1ppmNO 2 (seeFigure8). Figure6.(a)ResponsetimedependenceontheWEsurfacearea,(b)theresponsetimedependenceon theanalyteflowrate.Theerrorbarsrepresentstandarderrorofthemeanthatwascalculatedfrom threeconsecutiveexposurestothesameconcentration. Figure7.(a)RecoverytimedependenceontheWEsurfacearea,(b)recoverytimedependenceonthe analyteflowrate.TheerrorbarsrepresentStandarderrorofthemeanthatwascalculatedfromthree consecutiveexposurestothesameconcentration. Thus,ifthesensorwasexposedtothecalculateddetectionlimit(unitsofppbforsensorwiththe WEsurfaceareaof22.4mm 2 ,seeFigure9)anexpectedcurrentresponsewouldbewithintherange ofunitsofnanoampere(nA).Althoughsuchalowcurrentresponsecanbemeasured,thereisno guaranteewhethersuchasmallchangeinthemeasuredcurrentcannotbeattributedtosome
Sensors2020,20,10387of17 parasiticevents.Second,theactivationoftheelectrodesurfacemayrequireaspecificamountof analyteandsothecalculateddetectionlimitmaynaturallydifferfromtheactualvalue.Nevertheless, LODisindisputablyasensorparameterthatcontributestothedeterminationoftheoverallsensor performanceandisworthdetermining;especiallywhenwearenotprimarilyinterestedinthe absolutevalueofLODbutratherinitspotentialdependenceonWEsurfaceareaand/oranalyteflow rate(seeFigure9).WhiletheWEsurfaceareahadaminorimpactonLODforanalyteflowratesof 0.5,0.8and1L/min,asubstantialincreaseoftheLODvaluewasobservedforthesmallestWEarea (2.9mm 2 )atthelowestanalyteflowrate(0.1L/min).Thisfactwasprobablycausedbyacombination oflowsensitivityandnon‐ideal(non‐constant)backgroundcurrent,whichresultedinahighervalue ofthebackgroundcurrentnoise.AsimilarresultcanbeseeninFigure9b,wherenegligibleLOD dependenceonanalyteflowratewasobserved,withtheexceptionofthesensorwiththesmallest WEsurfaceareaatthelowestanalyteflowrate. Figure8.SensorresponsetostepwiseincreaseinNO 2 concentrationswithintherangeof0−1ppm(1step equaled100ppb),WEsurfaceareaof22.4mm 2 conditions:298K,40%RH,analyteflowrate1L/min. Figure9.(a)Limitofdetection(LOD)dependenceontheWEsurfacearea,(b)LODdependenceon theanalyteflowrate. Repeatabilitymayprovideinformationabouttheshort‐termstabilityofsensorresponse.This parameterwasdeterminedasaratioofthetriplestandarddeviationofthreeconsecutive measurements(seeFigure4)whenthetestedsensorwasexposedtothesameconcentrationleveland sensitivity.Figures10a,bshowrepeatabilitydependencesontheWEsurfaceareaandtheanalyte flowrate.Themeasurementofsensorrepeatabilitywasnaturallyinfluencedbytherepeatabilityof theactualconcentrationlevelinthetestchamberitself.Becausetheactualconcentrationlevelat
Sensors2020,20,10388of17 repeatedexposurescouldfluctuatewithintherangeoftensofppb(theworstvalueis±40ppbatflow rate0.5L/min,givenbytherepeatabilityoftheparticularflowthroughthemassflowcontroller),it isnecessarytotakethisfactintoaccountanddiscussonlycleargeneraltrends.Figure10aindicates changesofrepeatabilitywithoutacleartrendindependenceonWEsurfaceareafortheflowratesof 0.5,0.8and1L/min,whileconsiderablegrowingdependencewasobservedfortheflowrateof0.1 L/min,whichresultedinworserepeatabilityforalargerWEsurfacearea.Figure10bshowsrelatively minorchangesinsensorrepeatabilitywithintherangeof0.5−1L/min,whilesubstantiallyworsevalues ofrepeatabilitywereobservedforallWEsurfaceareaswhenanalyteflowratewasdecreasedto0.1 L/min.Theresponseofthesensorsshowedasignificantdependenceonthemutualorientationofthe sensorandtheanalyteflowdirection,aswellasonthelocationofthesensorinthetestchamber. Consideringpreviousmeasurements,furtherexperimentsregardingsensororientationandlocation werefocusedonlyonthesensorwiththelargestworkingelectrode. Figure10.(a)RepeatabilitydependenceontheWEsurfacearea,(b)repeatabilitydependenceonthe analyteflowrate. 3.2.EffectoftheSensorRotationonSensorParameters ThesecondsetofexperimentswasaimedatthesensorwiththelargestWEsurfacearea,where attentionwaspaidtotheeffectofthemutualorientationoftheWEsurfaceareaandtheanalyteflow directiononsensorparameters.Thesensorwasconnectedviasilverwiresandplacedapproximately 5mmabovethechamberbottom(seeFigure3).Weexposedthesensortoatestcycle(Figure4)under fourdifferentmutualorientationsofanalyteflowdirectionandsensorsurface(anglesof0°,45°,90° and270°,seeFigure11).Sensorsensitivitygraduallyincreasedasthesensorwasrotatedfromthe0° angleposition(analyteflowdirectionwasparalleltotheWEsurfacearea)tothe90°angleposition (analyteflowdirectionwasperpendiculartotheWEsurfacearea).Themutualangleof270°(the sensorwasfacingawayfromtheanalyteflowdirection)causedarapiddecreaseinsensitivity.An interestingfeaturewasobservedwhenthesensitivityofthesensorplacedatthechamberbottom(red pointinFigure11a,mutualanglenear0°isexpected)wascomparedwiththeresultsoftherotated sensorfromFigure11a.Itwasobservedthatthesensorpositioninzaxis,i.e.,theheightofthesensor abovethechamberbottom,wasimportantandalthoughbothsensitivitiesforthesepositionscould beexpectedtobesimilar,theydifferedmarkedly.Consideringthesefacts,itisobviousthatthe numberofdetectedgasmolecules,whichhitdirectlythesurfaceofWE,persecondincreaseswith theangle(from0°upto90°),thecurrentviasensorgrows(describedindetailintheSection3.3.)and thesensitivityishigher(seeFigure11).Thus,wecouldassumethatthesensitivityisrelatedtotheto thepressureexertedonsensor,i.e.,tofluidicconditionsaroundthesensor.Response/recoverytime dependenceisshowninFigure11b.Generally,responsetimeswerelowerthanrecoverytimesforall mutualorientationanglesbetweenanalyteflowandtheWEsurfacearea.Boththeresponse/recovery timeshowedaslightlydecreasingtrendasthemutualanglewasincreased.Whenthesensorwas facingawayfromtheanalyteflow(270°angle)bothtimesincreasedslightly.Theonlyexceptionwas observedwithrespecttotheaforementioneddescription.Thechangeofmutualanglefrom45°to90°
Sensors2020,20,10389of17 resultedinminorincreaseofthevalueofrecoverytimeincontrasttothecontinualdecreasingtrend ofresponsetime.Thisfactdoesnothaveaclearexplanationanditcouldbecausedbyspecificfluidic conditionsaroundthesensororattributedtoanotquitestablecurrentresponseofthesensoratthe angleof45°. Theimpactofsensorrotationonthelimitofdetectionandtherepeatabilityofsensorresponse isshowninFigure12.LODdecreasedastheangleofrotationincreased,excepttheanglevalueof 270°.BecauseLODwascalculatedastheratiooftriplestandarddeviationofbackgroundcurrent noiseandsensitivity,itcanbereasonablyexpectedthattheLODdependenceontherotationangle wouldfollowexactlytheoppositetrendtosensitivitydependenceinFigure11.Repeatabilitywas deterioratingwiththeincreasingangleofrotationfrom0°to90°.Whentheangleof270°wasset,the valueofrepeatabilitydecreasedtoalmosttheoriginvalueatthe0°angle. Figure11.(a)Sensitivitydependenceonthemutualorientationoftheanalyteflowdirectionandthe WEsurfacearea(22.4mm 2 );(b)response/recoverytimedependenceonthemutualorientationofthe analyteflowdirectionandtheWEsurfacearea(22.4mm 2 );conditions:298K,40%RH,analyteflow rate1L/min.TheerrorbarsinFigure11arepresentthestandarderroroftheslopeofthecalibration curve(sensitivity)thatwasconstructedbylinearregression.TheerrorbarsinFigure11brepresent standarderrorofthemeanthatwascalculatedfromthreeconsecutiveexposurestothesameconcentration. Figure12.(a)LODdependenceonthemutualorientationofanalyteflowdirectionandtheWE surfacearea(22.4mm 2 );(b)repeatabilitydependenceonmutualorientationofanalyteflowdirection andWEsurfacearea(22.4mm 2 );conditions:298K,40%RH,analyteflowrate1L/min. 3.3.TheEffectoftheSensorOrientationonDirectCurrent(DC)ResponseandCurrentFluctuationsunder EquilibriumConditions Thesignalresponsesofthelargestworkingelectrodewerefurtherinvestigatedunder equilibriumconditionswhenthesensorwasbeingkeptatparticularconditions(concentrationand
Sensors2020,20,103816of17 7. Barbri,N.E.;Duran,C.;Brezmes,J.;Cañellas,N.;Ramírez,J.L.;Bouchikhi,B.;Llobet,E.Selectivity EnhancementinMultisensorSystemsUsingFlowModulationTechniques.Sensors2008,8,7369–7379. 8. Martinelli,E.;Polese,D.;Catini,A.;D’Amico,A.;DiNatale,C.Self‐adaptedtemperaturemodulationin metal‐oxidesemiconductorgassensors.Sens.ActuatorsBChem.2012,161,534–541. 9. Kato,Y.;Yoshikawa,K.;Kitora,M.Temperature‐dependentdynamicresponseenablesthequalification andquantificationofgasesbyasinglesensor.Sens.ActuatorsBChem.1997,40,33–37. 10. Wachholz,F.;Biała,K.;Piekarz,M.;Flechsig,G.‐U.Temperaturepulsemodulatedamperometryatcompact electrochemicalsensors.Electrochem.Commun.2007,9,2346–2352. 11. Burgués,J.;Marco,S.LowPowerOperationofTemperature‐ModulatedMetalOxideSemiconductorGas Sensors.Sensors2018,18,339. 12. Macku,R.;Smulko,J.;Koktavy,P.;Trawka,M.;Sedlak,P.Analyticalfluctuationenhancedsensingby resistivegassensors.Sens.ActuatorsBChem.2015,213,390–396. 13. Smulko,J.M.;Trawka,M.;Granqvist,C.G.;Ionescu,R.;Annanouch,F.;Llobet,E.;Kish,L.B.New approachesforimprovingselectivityandsensitivityofresistivegassensors:Areview.Sens.Rev.2015,35, 340–347. 14. Chinh,N.D.;Quang,N.D.;Lee,H.;ThiHien,T.;Hieu,N.M.;Kim,D.;Kim,C.;Kim,D.NOgassensing kineticsatroomtemperatureunderUVlightirradiationofIn2O3nanostructures.Sci.Rep.2016,6,35066. 15. Kumar,R.;Goel,N.;Kumar,M.UV‐ActivatedMoS2BasedFastandReversibleNO2SensoratRoom Temperature.ACSSens.2017,2,1744–1752. 16. Caravati,E.M.;Anderson,K.T.BreathAlcoholAnalyzerMistakesMethanolPoisoningforAlcohol Intoxication.Ann.Emerg.Med.2010,55,198–200. 17. Jasinski,G.;Strzelczyk,A.;Koscinski,P.Gassamplingsystemformatrixofsemiconductorgassensors.IOP Conf.Ser.Mater.Sci.Eng.2016,104,012033. 18. Scandurra,G.;Cannatà,G.;Ciofi,C.Differentialultralownoiseamplifierforlowfrequencynoise measurements.Cit.AIPAdv.2011,1,022144. 19. Hashisaka,M.;Ota,T.;Yamagishi,M.;Fujisawa,T.;Muraki,K.Cross‐correlationmeasurementofquantum shotnoiseusinghomemadetransimpedanceamplifiers.Cit.Rev.Sci.Instrum.2014,85,054704. 20. Beulertz,G.;Geupel,A.;Moos,R.;Kubinski,D.J.;Visser,J.H.Accumulatinggassensorprinciple—Howto comefromconcentrationintegrationtorealamountmeasurements.Proc.Eng.2011,25,1109–1112. 21. Sedlák,P.;Kuberský,P.;Mívalt,F.Effectofvariousflowrateoncurrentfluctuationsofamperometricgas sensors.Sens.ActuatorsBChem.2019,283,321–328. 22. Kuberský,P.;Hamáček,A.;Nešpůrek,S.;Soukup,R.;Vik,R.Effectofthegeometryofaworkingelectrode onthebehaviorofaplanaramperometricNO2sensorbasedonsolidpolymerelectrolyte.Sens.ActuatorsB Chem.2013,187,546–552. 23. Kuberský,P.;Sedlák,P.;Hamáček,A.;Nešpůrek,S.;Kuparowitz,T.;Šikula,J.;Majzner,J.;Sedlaková,V.; Grmela,L.;Syrový,T.Quantitativefluctuation‐enhancedsensinginamperometricNO2sensors.Chem. Phys.2015,456,111–117. 24. Sedlak,P.;Kubersky,P.;Skarvada,P.;Hamacek,A.;Sedlakova,V.;Majzner,J.;Nespurek,S.;Sikula,J. Currentfluctuationmeasurementsofamperometricgassensorsconstructedwiththreedifferent technologyprocedures.Metrol.Meas.Syst.2016,23,531–543. 25. Kuberský,P.;Altšmíd,J.;Hamáček,A.;Nešpůrek,S.;Zmeškal,O.AnelectrochemicalNO2sensorbasedon ionicliquid:Influenceofthemorphologyofthepolymerelectrolyteonsensorsensitivity.Sensors2015,15, 28421–28434. 26. Kuberský,P.;Syrový,T.;Hamáček,A.;Nešpůrek,S.;Syrová,L.Towardsafullyprintedelectrochemical NO2sensoronaflexiblesubstrateusingionicliquidbasedpolymerelectrolyte.Sens.ActuatorsBChem. 2015,209,1084–1090. 27. Ahmadi,M.M.;Jullien,G.A.Current‐Mirror‐BasedPotentiostatsforThree‐ElectrodeAmperometric ElectrochemicalSensors.IEEETrans.CircuitsSyst.IRegul.Pap.2009,56,1339–1348. 28. Gášpár,R.;Soucek,S.Analysisofaheatexchangerforacogenerationunitusingcomputationalfluid dynamics.InProceedingsoftheASMEInternationalMechanicalEngineeringCongressandExposition, Proceedings(IMECE),Tampa,USA,3–9November2017;Volume6. 29. Durst,F.UnstableFlowsandLaminar‐TurbulentTransition.InFluidMechanics;Springer: Berlin/Heidelberg,Germany,2008;pp.495–522.
Sensors2020,20,103817of17 30. Hassibi,A.;Navid,R.;Dutton,R.W.;Lee,T.H.Comprehensivestudyofnoiseprocessesinelectrode electrolyteinterfaces.J.Appl.Phys.2004,96,1074–1082. 31. Fourcade,B.;Tremblay,A.‐M.S.DiffusionNoiseofFractalNetworksandPercolationClusters.Phys.Rev. B1986,34,7802. 32. Green,M.E.Diffusionand1/fnoise.J.Membr.Biol.1976,28,181–186. 33. Musha,T.;Higuchi,H.TrafficCurrentFluctuationandtheBurgersEquation.Jpn.J.Appl.Phys.1978,17, 811–816. 34. Roach,P.E.Thegenerationofnearlyisotropicturbulencebymeansofgrids.Int.J.HeatFluidFlow1987,8, 82–92. 35. Kubersky,P.;Sedlak,P.Amperometricgassensorindifferentorientationtowardsanalyteflowatdifferent flowratesanddifferentconcentrations.MendeleyData2020,doi:10.17632/4JHGXXKKT4.1. ©2020bytheauthors.LicenseeMDPI,Basel,Switzerland.Thisarticleisanopenaccess articledistributedunderthetermsandconditionsoftheCreativeCommonsAttribution (CCBY)license(http://creativecommons.org/licenses/by/4.0/).