sensors Article
Conductometric Soot Sensors: Internally Caused Thermophoresis as an Important Undesired Side Effect Gunter Hagen *, Christoph Spannbauer, Markus Feulner, Jaroslaw Kita, Andreas Müller and Ralf Moos Bayreuth Engine Research Center (BERC), University of Bayreuth, Department of Functional Materials, 95440 Bayreuth, Germany * Correspondence:
[email protected]; Tel.: +49-921-55-7406 Received: 17 September 2018; Accepted: 15 October 2018; Published: 19 October 2018
Abstract: Particulate matter sensors are of interest for application in the exhaust of any combustion processes, especially for automotive aftertreatment systems. Conductometric soot sensors have been serialized recently. They comprise planar interdigital electrodes (IDE) on an insulating substrate. Between the IDEs, a voltage is applied. Soot deposition is accelerated by the resulting electric field due to electrophoresis. With increasing soot deposition, the conductance between the IDE increases. The timely derivative of the conductance can serve as a sensor signal, being a function of the deposition rate. An increasing voltage between the IDE would be useful for detecting low particle exhausts. In the present study, the influence of the applied voltage and the sensor temperature on the soot deposition is investigated. It turned out that the maximum voltage is limited, since the soot film is heated by the resulting current. An internally caused thermophoresis that reduces the rate of soot deposition on the substrate follows. It reduces both the linearity of the response and the sensitivity. These findings may be helpful for the further development of conductometric soot sensors for automotive exhausts, probably also to determine real driving emissions of particulate matter. Keywords: particulate matter sensors; exhaust gas aftertreatment; particulate filter; soot deposition; electrophoresis; thermophoresis; dynamometer
1. Introduction Fine dust and particulate matter (PM) emissions from combustion processes, especially from diesel combustion, may cause serious health problems. The exhaust gas aftertreatment of diesel engines becomes increasingly complicated due to tightening emission standards [1–3]. The determination of vehicle emissions during defined driving cycles on chassis dynamometers is going to be replaced by real-driving emissions measurements [4]. Therefore, prospective needs will concern reliable on-board measurement systems—also for PM [5,6]. Particle abatement with diesel particulate filters (DPF) is indispensable in future [5,6]. Additionally, in the field of gasoline fueled passenger cars, the PM issue may become even more serious, since nano-sized soot emissions have to be avoided [7]. In the past, particle mass was the key target, but recently, a particulate number limit has also been introduced. Therefore, gasoline particulate filters (GPF) have also been serialized [8,9]. In addition to that, on-board-diagnostics (OBD) has to ensure permanently the correct functionality of all exhaust gas aftertreatment systems, which is here the filtering efficiency of particulate filters [10–14]. For PM aftertreatment, as the current state-of-the-art, ceramic wall-flow filters are applied in the exhaust pipe. Soot from the raw exhaust is trapped over a certain time span, but the filter has
Sensors 2018, 18, 3531; doi:10.3390/s18103531
www.mdpi.com/journal/sensors
Sensors 2018, 18, 3531
2 of 12
Sensors 2018, 18, x FOR PEER REVIEW
2 of 12
to be regenerated during operation to avoid clogging. Therefore, soot loadprediction prediction models models based regenerated during operation to avoid clogging. Therefore, soot load based on on differential pressure sensors signals and engine characteristic maps are in serial use to estimate differential pressure sensors signals and engine characteristic maps are in serial use to estimate the filter loading state The more the exactly filter loading is known, theefficiently more efficiently the can filterthe loading state [12]. The[12]. more exactly filterthe loading is known, the more the system system can be operated. Filter regeneration at higher exhaust gas temperatures to burn off the be operated. Filter regeneration at higher exhaust gas temperatures to burn off the trapped soot causes trapped soot causes higher fuel consumption and should be carried out only when necessary [15,16]. higher fuel consumption and should be carried out only when necessary [15,16]. A promising novel A promising novel approach uses microwave sensors to determine directly the filter loading degree in approach uses microwave sensors to determine directly the filter loading degree in operation [12,17–19]. operation [12,17–19]. For For OBD-purposes, soot tools. Several Severalprinciples principlesareare research OBD-purposes, sootsensors sensorsare are useful useful tools. in in thethe research or or development state [20–26]. Conductometric (sometimes also called “resistive”) soot sensors are development state [20–26]. Conductometric (sometimes also called “resistive”) soot sensors arevery oftenvery discussed, in industry [20,27–29] and academia [30–33]. A typical sensor is sketched in Figure often discussed, in industry [20,27–29] and academia [30–33]. A typical sensor is sketched in Figure 1. This device is explained detail, since is alsofor used the experiments described below. 1. This device is explained in detail,insince it is alsoit used thefor experiments described below.
Figure 1. Sketch of the sensor setup: The sensors front side comprises interdigital electrodes (IDE)
Figure 1. Sketch of the sensor setup: The sensors front side comprises interdigital electrodes (IDE) where soot deposits yield electrical conductive pathways over the spacing. On the reverse side, where soot deposits yield electrical conductive pathways over the spacing. On the reverse side, a foura four-wire thick-film heater allows temperature-controlled sensor regeneration as well as temperature wiremeasurement thick-film heater allows temperature-controlled sensor regeneration as well as temperature and temperature adjustment during soot deposition. measurement and temperature adjustment during soot deposition. On the upper side of an insulating, high temperature stable alumina substrate, interdigital electrodes (IDE) are screen-printed with Pt high paste temperature (LPA88-11S, Heraeus, Hanau, Germany), dried and On the upper side of an insulating, stable alumina substrate, interdigital fired. The electrode area is about 5 mm × 5 mm and comprises lines and spaces, both of 100 µm width. electrodes (IDE) are screen-printed with Pt paste (LPA88-11S, Heraeus, Hanau, Germany), dried and theelectrode reverse side of the substrate, a platinum thick-film structure is integrated. It isboth designed a way fired.On The area is about 5 mm × 5 mm and comprises lines and spaces, of 100inµm width. that the temperature distribution on the sensor side is homogenous, if one uses the platinum thick-film On the reverse side of the substrate, a platinum thick-film structure is integrated. It is designed in a structure as a heater [34]. Since it is designed in a special four-wire technique, the tip resistance can be way that the temperature distribution on the sensor side is homogenous, if one uses the platinum measured and the temperature can be deduced from the tip resistance. The (almost linear) correlation thick-film structure as a heater [34]. Since it is designed in a special four-wire technique, the tip function between heater resistance and sensor temperature is calibrated for each sensor in the lab resistance bethe measured andheater the temperature can betwo deduced from the tip resistance. before can using device. The structure provides different functionalities. Firstly, itThe can(almost be linear) correlation function between heater resistance and sensor temperature calibrated for used for heating, i.e., for generating a homogenous temperature distribution in theis IDE area either toeach ◦ sensor in the lab The heater structure provides twoC)different regenerate thebefore sensorusing surfacethe bydevice. oxidation of soot at high temperatures (600 or to heatfunctionalities. the sensor Firstly, be used for heating, i.e., for itgenerating a homogenous temperature distribution in the tip it to can a desired temperature. Secondly, enables measuring the sensor temperatures during soot deposition, is, in fact—if a very smallby heater currentof is soot applied—also a measurement(600 of the IDE area eitherwhich to regenerate the only sensor surface oxidation at high temperatures °C) or actual exhaust gastip temperature. All feed lines as well as the heateritstructure covered by the a glass to heat the sensor to a desired temperature. Secondly, enablesaremeasuring sensor ceramic passivation (QM 42, DuPont, CCI Eurolam, Dreieich, Germany) to avoid interconnections by temperatures during soot deposition, which is, in fact—if only a very small heater current is soot particles. More details on the sensor setup can be found in previous publications [35,36]. applied—also a measurement of the actual exhaust gas temperature. All feed lines as well as the Several modifications, especially for serial applications, have been reported in heater structure are covered by a glass ceramic passivation (QM 42, DuPont, CCI Eurolam, Dreieich, literature [20,26–29,31,32]. Not depending on the manufacturer or research lab, most of the Germany) to avoid interconnections by soot particles. More details on the sensor setup can be found conductometric soot sensors are operated similarly. By applying a voltage UIDE to the IDEs, in previous publications [35,36]. an electrical current I can be measured as soon as the deposited soot particles form a first electrical Several modifications, especially serial applications, beentime, reported in literaturesoot [20,26– conductive percolation path betweenfor these electrodes. After have this blind with proceeding 29,31,32]. Not depending on the manufacturer research most of the conductometric deposition, the current increases. In other words,orthe sensor islab, an integrating device following the soot sensors are operated By applying a voltage UIDE as to described the IDEs,inan electrical I can be dosimeter principlesimilarly. with an instantaneous electrical readout, [37]. During acurrent subsequent ◦ C to burn off the soot. When cooled down to regeneration, is heated to several hundred measured as soonthe assensor the deposited soot particles form a first electrical conductive percolation path ambient temperature, i.e., to exhaust a new measurement cycle can deposition, start, again beginning between these electrodes. After thistemperature, blind time, with proceeding soot the current with a blind time. In addition to a conductometric readout, capacitive measurements have also beenwith increases. In other words, the sensor is an integrating device following the dosimeter principle investigated [23,24]. A typical operation mode of a conductometric soot sensor is described in Figure 2. an instantaneous electrical readout, as described in [37]. During a subsequent regeneration, the sensor
is heated to several hundred °C to burn off the soot. When cooled down to ambient temperature, i.e., to exhaust temperature, a new measurement cycle can start, again beginning with a blind time. In addition to a conductometric readout, capacitive measurements have also been investigated [23,24]. A typical operation mode of a conductometric soot sensor is described in Figure 2.
Sensors 2018, 18, 3531 Sensors 2018, 18, x FOR PEER REVIEW
3 of 12 3 of 12
Figure 2. Typical shape of the sensor’s raw signal during one sensing cycle with increasing current I Figure 2. Typical shape of the sensor’s raw signal during one sensing cycle with increasing current I after a certain “blind” time before the percolation threshold is reached and a regeneration peak after after a certain “blind” time before the percolation threshold is reached and a regeneration peak after increasing the sensor’s temperature for soot burn off. The regeneration peak is due to the increased increasing the sensor’s temperature for soot burn off. The regeneration peak is due to the increased soot conductivity when heating up. The slope of the linear increasing current during soot deposition, soot conductivity when heating up. The slope of the linear increasing current during soot deposition, which is proportional to the conductance, often acts as the sensor signal. which is proportional to the conductance, often acts as the sensor signal.
In our former work, we already showed the possibility for soot mass flow detection with these In our former work, we already possibility for the sootsoot mass flow detection with these dosimeter-like conductometric sensors.showed A directthe relation between mass flow in the raw exhaust dosimeter-like conductometric A direct between the soot the raw of a diesel engine and the timelysensors. derivative of the relation current (dI/dt) between themass IDE flow whenin a constant exhaust of a diesel engine and the timely derivative of the current (dI/dt) between the IDE when a voltage is applied was found [35,36]. constant voltage is applied was found [35,36]. In several contributions, the influence of the applied voltage was already found to be a significant In several the influence of the applied voltage waswith already to be a significant parameter for contributions, soot deposition. Higher electrical fields interact the found charged soot particles. parameter for soot deposition. Higher electrical fields interact with the charged soot particles. Another option is that electrical charges might be induced by the electrical field. Anyway, Another the soot option is that electrical charges surface might beby induced by the electrical field. Anyway, the soot isisaccelerated is accelerated to the sensor’s the Coulomb forces and thus soot collection enhanced to the sensor’s surface by the However, Coulomb forces andto thus soot collection is enhanced electrophoretically [38–43]. in order reliably determine particulateelectrophoretically mass or particle [38–43]. However, in order to reliably determine particulate mass or particle number, a deeper number, a deeper understanding of the deposition mechanisms on the sensor surface is needed. understanding of thecomplex deposition mechanisms on the sensor surface is needed. Soot is a highly matter. Characteristics like morphology, carbon content, adhered volatile Soot is a highly complex matter. Characteristics like carbon adhered organic components content or water, or its electrical chargemorphology, will also influence thecontent, interaction with volatile organic components content or water, or its electrical charge will also influence the a measuring device [7,44–49]. For a further knowledge-based development of conductometric soot interaction a measuring devicevoltage [7,44–49]. a further knowledge-based of sensors, thewith influences of the applied andFor of the sensor temperature on thedevelopment soot deposition conductometric soot sensors, the influences of the applied voltage and of the sensor temperature on needs to be investigated. In order to keep flexibility with changing geometries, electrode designs, the soot deposition housings, needs to be investigated. to keepplanar flexibility with changing geometries, sensor orientations, or to even be ableIntoorder manufacture capacitive devices [24], we used electrode designs, sensor orientations, housings, or to even be able to manufacture planar capacitive our own sensor structure as shown in Figure 1. It is fully manufactured in thick-film technology, devicesis[24], used ownproduction sensor structure as shown in Figure It is fullythat manufactured incan thickwhich verywe close toour typical type sensors [27,28]. It is1.expected our findings be film technology, which is very close to typical production type sensors [27,28]. It is expected that our transferred to commercial sensors as well. findings can be transferred to commercial sensors as well. 2. Sensor Design and Experimental Setup 2. Sensor Design and Experimental Setup It is the intention of the present study to contribute to a better understanding of factors that It is the of the present study Therefore, to contribute to a conductometric better understanding factorswere that influence sootintention deposition on such sensors. simple sensorofdevices influence soot deposition on such sensors. Therefore, simple conductometric sensor were built as shown in Figure 1 and operated as explained above. In addition to that, in somedevices experiments, built as shownfilm in Figure 1 and operated as explained In addition to that, in tip, some experiments, the platinum structure is used to adjust a definedabove. temperature on the sensor i.e., on the IDE, the platinum film structure used to adjust a defined temperature on the higher sensorthan tip, i.e., the IDE, during soot deposition. Here,isthe sensor temperature is set to a few degrees the on exhaust gas during soot deposition. Here, the sensor temperature is setare to areported few degrees higher than theofexhaust gas temperature to study thermophoresis effects. The results in the second part this study. temperature to study effects. were The results are reported in the second of the thissensors study. After wiring andthermophoresis housing, the sensors mounted into the exhaust pipe. part Here, After wiring and housing, the sensors were mounted into the exhaust pipe. Here, the sensors either face the gas flow with its IDE area or are oriented in a way that the sensor is mounted exactly either face the gas flow with its IDEwith areathe or electrode are oriented inbeing a waybehind that thethe sensor is mounted exactly perpendicular to the exhaust flow area substrate in the shadow perpendicular to the exhaust flow with the electrode area being behind the substrate in the shadow zone. No protection caps were used for these experiments. The sensor current and other parameters zone. No protection caps were voltage used forbetween these experiments. The current andand other parameters like heater resistance, applied the IDE (UIDE ), sensor or heater voltage heater current like heater resistance, applied voltage between the IDE (UIDE), or heater voltage and heater current during active heating were recorded with a digital multimeter (2700 series, Keithley, Tektronix, Munich, Germany).
Sensors 2018, 18, 3531
4 of 12
during active heating were recorded with a digital multimeter (2700 series, Keithley, Tektronix, Munich, Germany). Experiments in the dynamometer test bench, i.e., in real exhausts, were conducted in the exhaust Sensors 2018, 18, x FOR PEER REVIEW 4 of 12 pipe of a 2.1 L diesel engine (Figure 3). To change the amount of soot during operation, the boost in the dynamometer real exhausts, were conducted in the exhaust pressure pboostExperiments was varied between 1.25 test barbench, and i.e., 1.15inbar. All other engine parameters were kept pipe of a 2.1 L diesel engine (Figure 3). To change the amount of soot during operation, the boost Commented [M1]: Should i constant (1000 rpm, 25% accelerator pedal position, injection pressure pinj = 550 bar). Most of the pressure pboost was varied between 1.25 bar and 1.15 bar. All other engine parameters were kept it. tests wereconstant conducted several times at different days to get an idea of the repeatability of the soot (1000 rpm, 25% accelerator pedal position, injection pressure pinj = 550 bar). Most of the tests generation and to distinguish scattering soot concentration variations) from signal were conducted several times at different(coming days to getfrom an idea of the repeatability of the soot generation Commented [RM2R1]: Don and to distinguish scattering (coming from soot concentration variations)Particle from signal changes changes (resulting from the influencing parameters to be investigated). concentration data (resulting the influencing to be measurement investigated). Particle concentration were were obtained by from a commercial sootparameters nanoparticle device (Pegasor)data simultaneously obtained by a commercial soot nanoparticle measurement device (Pegasor) simultaneously during during all experiments. all experiments.
Figure 3. Setup for real exhaust measurements with a 2.1 L diesel engine.
To investigate the effect of soot deposition, the sensors were operated under constant exhaust Figure 3. Setup for real exhaust measurements with a 2.1 L diesel engine. Commented [M3]: Please c conditions during the soot-collecting phase. A certain dc voltage was applied between the electrodes investigate the effect of soot deposition, the sensors wereregeneration operated under constant exhaust Commented [RM4R3]: Don (UIDE ) and theToresulting current I was recorded. After that, soot was initiated by heating conditions during the soot-collecting phase. A certain dc voltage was applied between the electrodes up the device. Regeneration leads to the well-known peak-shape in the current signal as shown in Commented [M5]: Please c (UIDE) and the resulting current I was recorded. After that, soot regeneration was initiated by heating Figure 2 (and also for instance in [12,27,36]) followed by a zero current phase, when all soot is burned up the device. Regeneration leads to the well-known peak-shape in the current signal as shown in Commented [RM6R5]: Don off. A newFigure soot2measurement cycleinis[12,27,36]) then started. Within each engine operation point, at least three (and also for instance followed by a zero current phase, when all soot is burned off.recorded A new sootunless measurement cycle stated is then started. each engine operation point, at least three cycles were otherwise in the Within respective section. cycles were recorded unless otherwise stated in the respective section. As a first step, dynamometer test at different engine operation points by varying the boost As a first step, dynamometer test at different engine operation points by varying the boost pressure to achieve different soot concentrations were conducted. One series included three sensor pressure to achieve different soot concentrations were conducted. One series included three sensor loading cycles, threeatdifferent sootsoot concentrations, respectively. measurement cycle, loadingatcycles, three different concentrations, respectively. ForFor eacheach singlesingle measurement cycle, the sensorthe signal as as a slope inmA/s mA/s (only the linear of the I(t)-curve sensordI/dt signal was dI/dtcalculated was calculated a slopevalue value in (only the linear part ofpart the I(t)-curve taken to calculate the slope, as indicated Figure 2). measurement seriesseries provided nine was takenwas to calculate the slope, as indicated ininFigure 2).So, So,one one measurement provided nine data points, three for each soot concentration. These series were repeated several times at different data points, three for each soot concentration. These series were repeated several times at different days, with the dc voltage applied between the sensor electrodes being varied from 20 to 60 V. days, with the dc voltage applied between the sensor electrodes being varied from 20 to 60 V. 3. Results and Discussion for Real Exhaust Measurements
3. Results and Discussion for Real Exhaust Measurements 3.1. Influence of the Applied Voltage/Electrophoresis
3.1. Influence of the Applied Voltage/Electrophoresis In Figure 4, three typical and representative single sensor loading cycles under almost equal soot 3) are shown. The voltage between the IDEs UIDE was varied. The data are concentrations mg/mand In Figure 4, three (17 typical representative single sensor loading cycles under almost equal soot shifted to a joint starting time t = 0, which is the end of the previous regeneration procedure. The raw 3 concentrations (17 mg/m ) are shown. The voltage between the IDEs UIDE was varied. The data data I(t) already show the described effect: a higher voltage leads to a higher slope of the current. At are shifted toglance, a joint starting time t this = 0,towhich is the endthat ofdrives the previous regeneration procedure. first one would attribute the higher voltage a higher current in agreement The raw data already show the a higher voltage leads to show a higher slope of the with I(t) Ohm’s law. However, thedescribed (calculated)effect: conductance curves (G = I/U the same IDE) still behavior: the higher voltage, the higher the to slope the conductance dG/dt.drives In other current. At first glance, onethe would attribute this theofhigher voltage that a words, higherthe current in soot deposition rate is in fact influenced by the applied voltage and one may assume that the amount agreement with Ohm’s law. However, the (calculated) conductance curves (G = I/UIDE ) still show of deposited soot on the sensor surface increases with the applied voltage. The data shown here are the same taken behavior: higher thewith voltage, thefacing higher slopebut ofthe the conductance dG/dt. with a the sensor oriented electrodes the the gas flow, findings are very similar for In other words, thethesoot deposition rate is in fact influenced by the applied voltage and one may assume that opposite mounting. the amount of deposited soot on the sensor surface increases with the applied voltage. The data shown here are taken with a sensor oriented with electrodes facing the gas flow, but the findings are very similar for the opposite mounting.
Sensors 2018, 18, x FOR PEER REVIEW Sensors 2018, 18, 3531 Sensors 2018, 18, x FOR PEER REVIEW
5 of 12 5 of 12 5 of 12
Figure 4. 4. Raw and calculated conductance data G (b) measurement cyclescycles (soot Figure Rawdata dataI(t) I(t)((a) ((a) and calculated conductance data G for (b) three for three measurement Figure 4. Raw data I(t) ((a) voltages and calculated conductanceregeneration) data G (b) forunder three measurement cycles (soot collection with different and subsequent similar conditions (soot collection with different voltages and subsequent regeneration) under similar conditions 3). collection with= 17 different voltages and subsequent regeneration) under similar (soot concentration mg/m The 3curves shifted a common that conditions is the that end is of the (soot concentration = 17 mg/m ). Theare curves aretoshifted to a starting commontime starting time the 3). The curves are shifted to a common starting time that is the end of the concentration = 17 mg/m prior regeneration phase. end of the prior regeneration phase. prior regeneration phase.
Thepercolation percolationtime time blind time) with the applied and supports the The (or(or blind time) also also variesvaries with the applied voltage voltage and supports the previous The percolation time (or blind time) also varies with the applied voltage and supports the previous findings. It can be determined directly theand/or current the and/or the conductance For findings. It can be determined directly from thefrom current conductance curves.curves. For lower previous findings. It can be time determined directly from the current and/or conductance lower applied voltages, the blinduntil timeauntil acurrent first current appears increases. It takes sFor for applied voltages, the blind first appears increases. Itthe takes nearlynearly 100curves. s 100 for 20 V lower applied voltages, the blind time until a first current appears increases. It takes nearly 100 s for 20 V and just 40 s for 60 V until the first current flows. Hence, higher voltages support the formation and just 40 s for 60 V until the first current flows. Hence, higher voltages support the formation of 20 V and just 40 s between forbetween 60 V until the first current flows. Hence, higher voltages support the formation of conducting paths electrodes. conducting paths thethe electrodes. of conducting betweenfield-supported the electrodes. soot The effect effectpaths of electrical electrical field-supported soot deposition deposition isis well wellknown. known. Soot Soot particles particles from from The of The effect of electrical field-supported soot deposition is well known. Soot particles from exhausts are electrically charged and are be attracted to the electrodes by a potential difference in exhausts charged and are be attracted to the electrodes by a potential difference exhausts are electrically charged and are be attracted to the electrodes by a potential difference in such setups [38–43]. are in such setups [38–43].Detailed Detailedinformation informationon onthe thecharging chargingof ofsoot soot particles particles and electrophoresis are such setups [38–43]. Detailed information on the charging of soot particles and electrophoresis are given in [50]. given in [50]. givenFigure in [50].55 shows Figure shows the the sensor sensor calibration calibration curves, curves, which which include include the the timely timely derivatives derivatives of of the the Figure shows sensor calibration curves, which the timely derivatives of the conductance5dG/dt dG/dt a measurement for point each pointinclude plotted over the soot concentration conductance ofofathe measurement cyclecycle for each plotted over the soot concentration measured 3). Monotonous conductance of adevice measurement cycle each3in plotted over thecorrelations soot concentration measured by dG/dt the Pegasor sensing device (values given mg/m between by the Pegasor sensing (values given infor mg/m ).point Monotonous correlations between dG/dt 3). Monotonous correlations between measured by the Pegasor sensing device (values given in mg/m dG/dt and soot concentrations arefor found both types of sensor orientation. Furthermore, the and soot concentrations are found both for types of sensor orientation. Furthermore, the provided dG/dt andconductance soot for types of sensor orientation. Furthermore, the provided data are are normalized to both atemperature common temperature of 300 °C. Higher voltages conductance dataconcentrations are normalized to found a common of 300 ◦ C. Higher voltages result in a provided conductance data are normalized to a common temperature of 300 °C. Higher voltages result in a higher sensor signal. All these findings agree with the literature [35]. higher sensor signal. All these findings agree with the literature [35]. result in a higher sensor signal. All these findings agree with the literature [35].
Figure soot concentration) concentration) for Figure 5. 5. Sensor Sensor calibration calibration curves curves (calculated (calculatedconductance conductancedG/dt dG/dt vs. vs. soot for different different types of sensor orientation in the exhaust, either IDE facing the gas flow (a) or in the opposite Figure 5. Sensor conductance dG/dtthe vs.gas sootflow concentration) formounting different types of sensor calibration orientationcurves in the(calculated exhaust, either IDE facing (a) or in the opposite position types of (b). sensor orientation in the exhaust, either IDE facing the gas flow (a) or in the opposite mounting position (b).
mounting position (b).
Sensors 2018, 18, x FOR PEER REVIEW Sensors 2018, 18, 3531
6 of 12 6 of 12
The normalization was carried out by a linear calibration curve of the temperature dependency of the conductance. Forwas thatcarried purpose, sensor was operatedcurve in theofexhaust and soot dependency was deposited The normalization outaby a linear calibration the temperature of for 150 s at U = 60For V. This typicalwas sootoperated layer on in thethe electrodes. Thesoot sensor then taken the conductance. that produced purpose, aasensor exhaust and waswas deposited for out sofatthe exhaust. Using the internal heater, between 150 °Csensor and 375 °Cthen weretaken adjusted 150 U= 60 V. This produced a typical soottemperatures layer on the electrodes. The was out ◦ C and 375 In ◦ C all and conductances measured for 20, 40, and 60 V,150 respectively. experiments, of thethe exhaust. Using thewere internal heater, temperatures between were adjusted anda temperature dependency of dG/dT ca.40, 0.3and µS/°C was found. the conductances were measured forof20, 60 V, respectively. In all experiments, a temperature ◦ To obtain higher of sensor a reduced percolation time, electrophoresis has to be dependency of dG/dT ca. 0.3signals µS/ C and was found. increased by ahigher higher voltage. This an interesting phenomenon that ishas described in the To obtain sensor signals andcauses a reduced percolation time, electrophoresis to be increased following section. by a higher voltage. This causes an interesting phenomenon that is described in the following section. 3.2. Influence of Thermophoresis In Figure Figure 6a 6a (upper (upper graph), graph), raw raw data data of of the the conductance conductance for for three three measurement measurement cycles cycles with with subsequent regeneration regeneration for for applied appliedvoltages voltagesbetween betweenthe theinterdigital interdigital electrodes 20are V electrodes of of 60,60, 40, 40, 20 V are shown. engine operation point and, therefore,both boththe theexhaust exhausttemperature temperature and and the soot shown. TheThe engine operation point and, therefore, soot 3 concentration during thethe experiment. According to thermocouple data, concentration(30 (30mg/m mg/m3))were werekept keptconstant constant during experiment. According to thermocouple the exhaust temperature variedvaried only in a range of ±5 of K.±5 The part of theofconductance is of data, the exhaust temperature only in a range K.increasing The increasing part the conductance special interest here (area highlighted by a dashed ellipse). As already known from Figure 3, the slope is of special interest here (area highlighted by a dashed ellipse). As already known from Figure 3, the of the of conductance curve curve increases with increasing voltages and first paths occur slope the conductance increases with increasing voltages andpercolation first percolation paths faster. occur Astonishingly, for 60 V,for the at least at in least the second of the deposition period, faster. Astonishingly, 60slope V, thedecreases, slope decreases, in the half second halfsoot of the soot deposition whereas for 40 V for and40for V for the20 slope remains constant.constant. period, whereas V 20 and V the slope remains
Figure 6. Conductance over time during soot deposition for three cycles with different voltages Figure 6. Conductance over time during soot deposition for three cycles with different voltages UIDE UIDE between the sensing electrodes (a) and corresponding sensor temperature determined by the between the sensing electrodes (a) and corresponding sensor temperature determined by the reverse reverse side platinum meander element (b). The engine operation point was kept constant during the side platinum meander element (b). The engine operation point was kept constant during the experiment (soot concentration = 30 mg/m3 ). experiment (soot concentration = 30 mg/m3).
The respective temperature of the sensor element, which was measured by the internal heater The during respective of theissensor element, which measured by the internal heater structure the temperature whole experiment, plotted in Figure 6b. Inwas general, this temperature represents structure during the whole experiment, is plotted in Figure 6b. In general, this temperature represents the exhaust temperature. Considering the temperature at 50 s, i.e., after sufficient cooling from the exhaust temperature. thethe temperature at 50 s, i.e., sufficient cooling regeneration temperature, Considering one finds that operation conditions areafter relatively constant andfrom the regeneration temperature, one finds that the operation conditions are relatively constant and the ◦ exhaust temperature varies only in the range of 5 C. The eye-catching result in this diagram is the exhaust temperature varies only in the range of 5 °C. The eye-catching result in this diagram is ◦ increase of the sensor temperature during soot deposition for 60 V of about 19 C. When 40the V increase of the sensor temperature during soot deposition for 60 V of about 19 °C. When 40 V are ◦ are applied for soot deposition, the sensor temperature increases a little less but still by 9 C. applied for soot deposition, the sensor temperature increases a little less but still by 9 °C. No
Sensors 2018, 18, x FOR PEER REVIEW Sensors 2018, 18, 3531
7 of 12 7 of 12
temperature increase can be seen for UIDE = 20 V. This suggests that a temperature increase of the sensor is responsible for the slope of the in Figure No temperature increase candecreasing be seen for UIDE = 20conductance V. This suggests that6a. a temperature increase a temperature the exhaust gasofcan ruled out, an heating effect must of theSince sensor is responsibleincrease for the of decreasing slope thebe conductance in internal Figure 6a. be responsible for the observed behavior. If one calculates the power P IDE (P = U IDE·I = U IDE2·G) that Since a temperature increase of the exhaust gas can be ruled out, an internal heating effect must 2 generates Joule’s heat between the sensing electrodes and plots it vs the observed temperature be responsible for the observed behavior. If one calculates the power PIDE (P = UIDE ·I = UIDE ·G) increase, one obtains 7 (for Uthe IDE = 60 V). In other words, during soot the power P, that generates Joule’s Figure heat between sensing electrodes and plots it vs thedeposition, observed temperature which generates heat directly in the deposited soot, increases. At the beginning, the conductance G increase, one obtains Figure 7 (for UIDE = 60 V). In other words, during soot deposition, the power P, is zerogenerates since no current flows.in Since there is no soot, additional power there is no temperature increase. which heat directly the deposited increases. AtP, the beginning, the conductance G The higher the soot loading between the IDE, the higher is the conductance and the more heat is is zero since no current flows. Since there is no additional power P, there is no temperature increase. generated. At the end after 250 s, a temperature increase of 19 K occurs. According to Figure 7, The higher the soot loading between the IDE, the higher is the conductance and the more heat is applying 60At V the between the electrodes leads then to a heat generation of 0.6According W. If we compare generated. end after 250 s, a temperature increase of 19 K occurs. to Figurethat 7, with our experiences concerning the heater structure, materials data and simulations on similar applying 60 V between the electrodes leads then to a heat generation of 0.6 W. If we compare that sensor [51], an increase of K seems plausible. As adata conclusion, the temperature with oursetups experiences concerning the19 heater structure, materials and simulations on similarincrease sensor with higher applied voltages a direct and an origin.the Due to the higher voltage, soot setups [51], an increase of 19 K has seems plausible. Asindirect a conclusion, temperature increase withthe higher deposition rate is higher due to the electrophoretic effect. This causes thicker soot layers that lead to applied voltages has a direct and an indirect origin. Due to the higher voltage, the soot deposition a higher conductance of electrophoretic the soot films. In addition, generated quadratically with rate is higher due to the effect. This the causes thicker heat soot increases layers that lead to a higher the voltage. conductance of the soot films. In addition, the generated heat increases quadratically with the voltage.
Figure Figure 7.7. Internal Internal self-heating self-heating of of the the device: device: Correlation Correlation of of the the increase increase of of the thesensor sensortemperature temperature (determined from the resistance of the platinum structure) and the calculated electrical (determined from the resistance of the platinum structure) and the calculated electricalpower poweron onthe the sensing electrodes during soot deposition when 60 V were applied, calculated from the data in Figure 6. sensing electrodes during soot deposition when 60 V were applied, calculated from the data in Figure 6.
Therefore, we conclude that the internally generated heat reduces the rate of particulate Therefore, we conclude that the internally generated heat reduces the rate of particulate matter matter deposition on the sensor surface due to thermophoresis [26,52,53], in other words, internally deposition on the sensor surface due to thermophoresis [26,52,53], in other words, internally generated thermophoresis counteracts electrophoresis. This would explain the decreasing slope in the generated thermophoresis counteracts electrophoresis. This would explain the decreasing slope in conductance curve as it has been found out in Figure 5a for 60 V. the conductance curve as it has been found out in Figure 5a for 60 V. Another experiment was conducted to investigate this “internally caused” thermophoresis effect Another experiment was conducted to investigate this “internally caused” thermophoresis effect further. Again, a sensor was constantly operated in the indirect mode with the heater structure further. Again, a sensor was constantly operated in the indirect mode with the heater structure facing facing the gas flow, now with 40 V applied between the electrodes. The engine was operated the gas flow, now with 40 V applied between the electrodes. The engine was operated under constant under constant speed and load. We conducted several soot deposition cycles, one after the other. speed and load. We conducted several soot deposition cycles, one after the other. After each After each regeneration procedure, the heater was not completely switched off, but set to a certain regeneration procedure, the heater was not completely switched off, but set to a certain temperature. temperature. As a result, the sensor temperature (Tsensor ) during the soot deposition phase could be As a result, the sensor temperature (Tsensor) during the soot deposition phase could be deliberately deliberately increased by small values, dT, compared to the exhaust temperature (Texhaust ). The sensor increased by small values, dT, compared to the exhaust temperature (Texhaust). The sensor temperature temperature was calculated from the four-wire resistance of the internal platinum structure. In the was calculated from the four-wire resistance of the internal platinum structure. In the first cycle, no first cycle, no additional heating was applied (dT = Tsensor − Texhaust = 0 K). In the following cycles, additional heating was applied (dT = Tsensor − Texhaust = 0 K). In the following cycles, the temperature the temperature was increased stepwise to dT = 6, 11, 17, 24, and 33 K. The results are shown in was increased stepwise to dT = 6, 11, 17, 24, and 33 K. The results are shown in Figure 8. Figure 8.
Sensors 2018, 18, 3531 Sensors 2018, 18, x FOR PEER REVIEW
8 of 12 8 of 12
Figure 8. Conductance G for several soot deposition cycles at UIDE = 40 V during constant engine Figure 8. Conductance G for several soot deposition cycles at UIDE = 40 V during constant engine operation. After regeneration, an additional small power was applied to the sensor to establish a operation. After regeneration, an additional small power was applied to the sensor to establish a slightly higher temperature of the sensor tip compared to the exhaust temperature. This temperature slightly higher temperature of the sensor tip compared to the exhaust temperature. This temperature difference dT = Tsensor − Texhaust that was kept constant during each soot deposition phase. difference dT = Tsensor − Texhaust that was kept constant during each soot deposition phase.
It turned out that the higher the sensor temperature is, the longer it takes until the first percolation It and turned out that the higher sensor temperature is, the it takes until the first sets in the lower the slope of the the conductance curve is during sootlonger deposition. percolation sets in and the lower the slope of the conductance curve is during soot deposition. In Figure 9a, the slopes of the conductance dG/dt, as they were derived from Figure 8, are plotted Figure 9a, the slopes of the dT. conductance dG/dt, werefor derived from Figure 8, are plotted over In the temperature difference Since dG/dt is as a they measure the amount of soot that gets over the temperature difference dT. Since dG/dt is a measure for the amount of soot that gets deposited on the sensor per time interval, the monotonously decreasing curve supports the assumption deposited on the sensor per time interval, the monotonously decreasing curve supports the that “internally caused” thermophoresis is an effect that has to be considered when deducing soot assumption that “internally thermophoresis is an effect that has to be considered when concentrations from the slopecaused” of the conductance dG/dt. deducing soot concentrations from the slope of the conductance dG/dt. In addition to that, this assumption was verified by another analysis. During regeneration, In addition to that, this assumption was verified by another analysis. During the current and correspondingly the conductance shows a “peak” behavior. Theregeneration, step changethe in ◦ current and correspondingly the conductance shows a “peak” behavior. The step change in sensor sensor temperature up to 600 C leads to an almost sudden conductance increase due to the enhanced temperature up to 600 °C leads almostdecline sudden conductance due toand the enhanced soot soot conductivity, followed by to anan abrupt when the soot increase gets oxidized percolation is conductivity,The followed by an reaches abrupt zero decline when the soot sootisgets oxidized andbehavior percolation is interrupted. conductance as soon as the burned off. This is well interrupted. The conductance reaches zero as soon as the soot is burned off. This behavior is well known and often described in literature [12,27,36]. The area below the peak of the conductance curve known regeneration and often described literatureto[12,27,36]. Theofarea thebeen peakdeposited of the conductance curve during may be in correlated the amount sootbelow that has right before the during regeneration correlated to the amount of sootas that has deposited rightdepend before the regeneration starts as may both be effects, the increasing conductance well asbeen the declining curve, on regeneration starts as both effects, the increasing conductance as well as the declining curve, depend the thickness of the soot layer or its mass, respectively. In case of less deposited soot, the increasing on the of the soot layerup or its mass, respectively. In case of less deposited soot, the increasing part of thickness the peak during heating is smaller (less influence of temperature dependent conductivity part of theFurthermore, peak during heating up is smaller influence temperature dependent increase). the declining part of (less the peak startsoffaster (less amount will beconductivity burned off increase). Furthermore, the declining part of the peak starts faster (less amount will be burned off earlier). Therefore, the peak area is small. Analogously, a higher amount of soot causes a higher earlier).increase, Therefore, peak area istakes small. Analogously, a higher ofissoot causes higher current thethe total oxidation longer and as a result, the amount peak area larger. So, ita should current increase, the total oxidation takes longer and as a result, the peak area is larger. So, it should be possible to estimate the final soot loading (i.e., the amount of deposited soot mass) on the sensor be possible to estimate the final soot loading the amount of deposited on the sensor surface before regeneration by evaluation of(i.e., the peak area. We now plot soot thesemass) integral values of surface before regeneration by evaluation of vs. the the peaktemperature area. We now plot these integral of the the conductance curve during regeneration difference during sootvalues deposition. conductance during regeneration vs.applied the temperature during soot deposition. Since Since not onlycurve the parameters Tsensor or the voltage Udifference IDE determine the amount of soot on the not only the parameters T sensor or the applied voltage UIDE determine the amount of soot on the sensor sensor surface, but also the individual time intervals for soot deposition for each single measurement, surface, also the individual time intervals deposition for of each single measurement, we we have but to normalize the integral value resultsfor to asoot similar time span soot deposition to compare have to normalize the integral value results to a similar time span of soot deposition to compare different measurement/regeneration cycles. This is done by the linear relation of conductance over different measurement/regeneration done the linear relation ofof conductance over time for each measurement cycle. Allcycles. valuesThis referisnow toby a soot loading interval 220 s. Figure 9b time for each measurement cycle. All values refer now to a soot loading interval of 220 s. Figure 9b shows the correlation of the normalized integral values vs. dT. Here also, the exponential decrease shows the correlation of the normalized integral values vs. dT. Here also, the exponential decrease occurs with very similar parameters as found before (exp(−0.3x)). Therefore, the evaluation of the occursarea with very similar parameters found before (exp(−0.3x)). Therefore, the evaluation of the peak during regeneration can beasused as an additional indicator of the amount of deposited peak area during regeneration can be used as an additional indicator of the amount of deposited soot soot right before regeneration. This method helps to understand thermophoresis effects when using right before regeneration. This method helps to understand thermophoresis effects when using synthetic soot (e.g., from CAST devices). Here, the conductivity of the soot does not linearly increase
Sensors 2018, 18, 3531
9 of 12
Sensors 2018, 18, x FOR PEER REVIEW
9 of 12
synthetic soot (e.g., from CAST devices). Here, the conductivity of the soot does not linearly increase with temperature. temperature. Therefore, sufficient to to describe describe the the amount amount of of with Therefore, the the conductance conductance data data are are not not sufficient deposited soot [54]. deposited soot [54].
Figure 9. Slope of the conductance dG/dt (evaluated from the data in Figure 8) vs. temperature Figure 9. Slope of the conductance dG/dt (evaluated from the data in Figure 8) vs. temperature difference between sensor and exhaust gas dT (a). The normalized integral values of the conductance difference between sensor and exhaust gas dT (a). The normalized integral values of the conductance curve during regeneration show similar behavior (b). The inset shows the calculation for the integral curve during regeneration show similar behavior (b). The inset shows the calculation for the integral value exemplarily. value exemplarily.
In conclusion, thermophoresis strongly affects the rate of soot deposition on the sensor In conclusion, strongly between affects the rate of soot deposition thedeposition sensor device. device. Whereas a thermophoresis higher voltage applied the electrodes enhances on soot due Whereas a higher voltage applied between the electrodes enhances soot deposition due to to electrophoresis, a higher sensor temperature (which might be also caused by the applied high electrophoresis, higher sensorfavors temperature (which might be also caused by the applied high voltages betweenathe electrodes) thermophoresis. voltages between the electrodes) favors thermophoresis. This effect might also be practically used. It should be possible to “adjust” the measurement This effect might Very also be used. It should be possible toallow “adjust” the measurement ranges of soot sensors. finepractically electrode structures, for example, which detection of ultra-low ranges of soot sensors. Very fine electrode structures, for example, which allow detection ultrasoot concentrations (easy and fast path building) might be operated at higher temperatures inofcase of low soot concentrations (easy and fast path building) might be operated at higher temperatures in atmospheres with high amounts of soot. case of atmospheres with high amounts of soot. 4. Conclusions and Outlook 4. Conclusions and Outlook The present contribution considered the effect of different voltages that generate different electrical present the contribution considered the effect of different voltageson that different fieldsThe to support attraction of soot particles to enhance the deposition thegenerate surface of planar electrical fields to support the attraction of soot particles to enhance the deposition on the surface of conductometric soot sensors. In contrast to other studies, we carefully investigated the effect of planar conductometric sootdue sensors. In contrast other studies, we carefully the effect internal heating of the soot to higher applied to voltages. The internally causedinvestigated thermophoresis is an of internal heating of the soot due to higher applied voltages. The internally caused thermophoresis effect that hinders the application of higher voltages between the electrodes in conductometric soot is an effect that hinders the application of higherasvoltages between theconditions; electrodesthe in conductometric sensors. Both mechanisms occur in real exhausts well as in synthetic results coincide soot sensors. Both mechanisms occur in real exhausts well as in which synthetic conditions; the results in general. However, using synthetic soot causes someaschallenges should be discussed in a coincide in general. However, using synthetic soot causes some challenges which should be discussed separate paper. in a separate paper. these parameters might be used to influence directly and knowledge-based the Advantageously, Advantageously, these parameters might be used to influence directly andeven knowledge-based soot deposition or interaction with a sensor device. This could help to adjust the sensitivitythe of soot deposition or interaction with a sensor device. This could help to adjust even the sensitivity of soot sensors and its use in different atmospheres with varying soot concentrations. soot sensors and its use in different atmospheres with varying soot concentrations. As a future goal, we intend to develop soot sensors that are sensitive to certain soot fractions. Therefore, ongoing work will focus on lab measurements with defined particle fractions. Variations
Sensors 2018, 18, 3531
10 of 12
As a future goal, we intend to develop soot sensors that are sensitive to certain soot fractions. Therefore, ongoing work will focus on lab measurements with defined particle fractions. Variations will be necessary also concerning the IDE geometry or its layout, the mounting position, or even the sensor principle. The possibilities of influencing soot adsorption by the device’s operation strategy (voltage and temperature) might be adaptable to “select” defined soot fractions, e.g., to discriminate between different soot fractions. Here, the exhaust gas temperature and the exhaust velocity should be taken into account. Recently published investigations on modeling particle deposition mechanisms [55] indicate an influence of these parameters on the question of whether electrophoresis or thermophoresis plays the major role. Author Contributions: G.H. and R.M. designed the experiments; C.S., M.F. and A.M. performed the experiments; C.S. analyzed and evaluated the data; J.K. manufactured the sensor devices; G.H. and R.M. wrote the paper. Acknowledgments: The publication of this paper was funded by the German Research Foundation (DFG) and the University of Bayreuth in the funding program “Open Access Publishing”. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2. 3. 4. 5. 6.
7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
Maricq, M. Chemical characterization of particulate emissions from diesel engines: A review. J. Aerosol Sci. 2007, 38, 1079–1118. [CrossRef] Twigg, M.V.; Phillips, P.R. Cleaning the Air We Breathe—Controlling Diesel Particulate Emissions from Passenger Cars. Platin. Met. Rev. 2009, 53, 27–34. [CrossRef] Alkemade, U.; Schumann, B. Engines and Exhaust after Treatment Systems for Future Automotive Applications. Solid State Ion. 2006, 177, 2291–2296. [CrossRef] Franco, V.; Kousoulidou, M.; Muntean, M.; Ntziachristos, L.; Hausberger, S.; Dilara, P. Road vehicle emission factors development: A review. Atmos. Environ. 2013, 70, 84–97. [CrossRef] Burtscher, H.; Majewski, W.A.; Jääskeläinen, H. PM Measurement: In-Situ Methods. DieselNet 2014, 7, 1–25. Mohr, M.; Lehmann, U.; Rütter, J. Comparison of Mass-Based and Non-Mass-Based Particle Measurement Systems for Ultra-Low Emissions from Automotive Sources. Environ. Sci. Technol. 2005, 39, 2229–2238. [CrossRef] [PubMed] Harris, S.J.; Maricq, M. Signature Size Distributions for Diesel and Gasoline Engine Exhaust Particulate Matter. J. Aerosol Sci. 2001, 32, 749–764. [CrossRef] Kittelson, D.B. Engines and Nanoparticles: A Review. J. Aerosol Sci. 1998, 29, 575–588. [CrossRef] Guan, B.; Reggie, Z.; He, L.; Huang, Z. Review of the state-of-the-art of exhaust particulate filter technology in internal combustion engines. J. Environ. Manag. 2015, 154, 225–258. [CrossRef] [PubMed] Weigl, M.; Roduner, C.; Lauer, T. Particulate Filter Onboard Diagnostics by Means of a Particulate Sensor. In Proceedings of the FISITA 2010 World Automotive Congress, Budapest, Hungary, 30 May–4 June 2010. Brunel, O.; Duault, F.; Lavy, J.; Creff, Y.; Youssef, B. Smart Soot Sensor for Particulate Filter OBD. SAE Int. J. Passeng. Cars Electron. Electr. Syst. 2013, 6, 307–327. [CrossRef] Kamimoto, T. A review of soot sensors considered for on-board diagnostics application. Int. J. Eng. Res. 2016, 18, 631–641. [CrossRef] Uchiyama, T.; Fujie, H.; Aso, M. DPF Failure Detection Method and DPF Failure Detection Device. U.S. Patent 8,770,016 B2, 20 December 2012. DieselNet, On-Board Diagnostics. Available online: https://www.dieselnet.com/standards/us/obd.php (accessed on 21 February 2018). Johnson, T. Review of Diesel Emissions and Control. SAE Int. J. Fuels Lubr. 2010, 3, 16–29. [CrossRef] Rose, D.; Boger, T. Different Approaches to Soot Estimation as Key Requirement for DPF Applications. SAE Tech. Pap. 2009, 2009. [CrossRef] Sappok, A.; Bromberg, L. Loading and Regeneration Analysis of a Diesel Particulate Filter with a Radio Frequency-Based Sensor. SAE Tech. Pap. 2010, 2010. [CrossRef] Feulner, M.; Hagen, G.; Piontkowski, A.; Müller, A.; Fischerauer, G.; Brüggemann, D.; Moos, R. In-Operation Monitoring of the Soot Load of Diesel Particulate Filters—Initial Tests. Top. Catal. 2013, 56, 483–488. [CrossRef]
Sensors 2018, 18, 3531
19. 20. 21. 22. 23. 24. 25.
26.
27. 28. 29. 30. 31.
32. 33. 34.
35. 36.
37. 38. 39. 40. 41. 42.
11 of 12
Feulner, M.; Hagen, G.; Hottner, K.; Redel, S.; Müller, A.; Moos, R. Comparative Study of Different Methods for Soot Sensing and Filter Monitoring in Diesel Exhausts. Sensors 2017, 17, 400. [CrossRef] [PubMed] Husted, H.; Roth, G.; Nelson, S.; Hocken, L.; Fulks, G.; Racine, D. Sensing of Particulate Matter for on-Board Diagnosis of Particulate Filters. SAE Int. J. Eng. 2012, 5, 235–247. [CrossRef] Rostedt, A.; Marjamäki, M.; Yli-Ojanperä, J.; Keskinen, J.; Janka, K.; Niemelä, V.; Ukkonen, A. Non-Collecting Electrical Sensor for Particle Concentration Measurement. Aerosol Air Q. Res. 2009, 9, 470–477. [CrossRef] Reiß, S. Continental ePM sensor for detecting particle emission. In Proceedings of the 3rd International Conference Sensors for Exhaust Gas Cleaning and CO2 Reduction, Leipzig, Germany, 29–30 June 2016. Kondo, A.; Yokoi, S.; Sakurai, T.; Nishikawa, S.; Egami, T.; Tokuda, M.; Sakuma, T. New Particulate Matter Sensor for On Board Diagnosis. SAE Int. J. Eng. 2011, 4, 117–125. [CrossRef] Hagen, G.; Feulner, M.; Werner, R.; Schubert, M.; Müller, A.; Rieß, G.; Brüggemann, D.; Moos, R. Capacitive soot sensor for diesel exhausts. Sens. Actuators B Chem. 2016, 236, 1020–1027. [CrossRef] Wasisto, H.S.; Merzsch, S.; Uhde, E.; Waag, A.; Peiner, E. Partially Integrated Cantilever-Based Airborne Nanoparticle Detector for Continuous Carbon Aerosol Mass Concentration Monitoring. J. Sens. Sens. Syst. 2015, 4, 111–123. [CrossRef] Malik, A.; Abdulhamid, H.; Pagels, J.; Rissler, J.; Lindskog, M.; Nilsson, P.; Bjorklund, R.; Jozsa, P.; Visser, J.; Spetz, A.; et al. A Potential Soot Mass Determination Method from Resistivity Measurement of Thermophoretically Deposited Soot. Aerosol Sci. Technol. 2011, 45, 284–294. [CrossRef] Masoudi, M.; Sappok, A.G. Soot (PM) Sensors. DieselNet 2014, 7, 1–14. Ochs, T.; Schittenhelm, H.; Genssle, A.; Kamp, B. Particulate Matter Sensor for On Board Diagnostics (OBD) of Diesel Particulate Filters (DPF). SAE Int. J. Fuels Lubr. 2010, 3, 61–69. [CrossRef] Bartscherer, P.; Moos, R. Improvement of the Sensitivity of a Conductometric Soot Sensor by Adding a Conductive Cover Layer. J. Sens. Sens. Syst. 2013, 2, 95–102. [CrossRef] Hagen, G.; Feistkorn, C.; Wiegärtner, S.; Heinrich, A.; Brüggemann, D.; Moos, R. Conductometric Soot Sensor for Automotive Exhausts: Initial Studies. Sensors 2010, 10, 1589–1598. [CrossRef] [PubMed] Lloyd Spetz, A.; Huotari, J.; Bur, C.; Bjorklund, R.; Lappalainen, J.; Jantunen, H.; Schütze, A.; Andersson, M. Chemical sensor systems for emission control from combustions. Sens. Actuators B Chem. 2013, 187, 184–190. [CrossRef] Grob, B.; Schmid, J.; Ivleva, N.P.; Niessner, R. Conductivity for Soot Sensing: Possibilities and Limitations. Anal. Chem. 2012, 84, 3586–3592. [CrossRef] [PubMed] Fleischer, M.; Pohle, R.; Wiesner, K.; Meixner, H. Soot sensor for exhaust gase. In Proceedings of the Eurosensors XIX, Barcelona, Spain, 11–14 September 2005. Wiegärtner, S.; Hagen, G.; Kita, J.; Reitmeier, W.; Hien, M.; Grass, P.; Moos, R. Thermoelectric hydrocarbon sensor in thick-film technology for on-board-diagnostics of a diesel oxidation catalyst. Sens. Actuators B Chem. 2015, 214, 234–240. [CrossRef] Hagen, G.; Müller, A.; Feulner, M.; Schott, A.; Zöllner, C.; Brüggemann, D.; Moos, R. Determination of the Soot Mass by Conductometric Soot Sensors. Procedia Eng. 2014, 87, 244–247. [CrossRef] Feulner, M.; Hagen, G.; Müller, A.; Schott, A.; Zöllner, C.; Brüggemann, D.; Moos, R. Conductometric Sensor for Soot Mass Flow Detection in Exhausts of Internal Combustion Engines. Sensors 2015, 15, 28796–28806. [CrossRef] [PubMed] Marr, I.; Groß, A.; Moos, R. Overview on Conductometric Solid-State Gas Dosimeters. J. Sens. Sens. Syst. 2014, 3, 29–46. [CrossRef] Grondin, D.; Breuil, P.; Viricelle, J.P.; Vernoux, P. Development of a particulate matter sensor for diesel engine. Procedia Eng. 2015, 120, 1237–1240. [CrossRef] Grondin, D.; Geara, S.; Breuil, P.; Viricelle, J.P.; Vernoux, P. Influence of Electrodes Polarization on the Response of Resistive Soot Sensor. Procedia Eng. 2016, 168, 31–34. [CrossRef] Grondin, D.; Westermann, A.; Breuil, P.; Viricelle, J.P.; Vernoux, P. Influence of key parameters on the response of a resistive soot sensor. Sens. Actuators B Chem. 2016, 236, 1036–1043. [CrossRef] Kittelson, D.B.; Pui, D.Y.H.; Moon, K.C. Electrostatic Collection of Diesel Particles. SAE Tech. Pap. 1986, 19–30. [CrossRef] Maricq, M. On the electrical charge of motor vehicle exhaust particles. Aerosol Sci. 2006, 37, 858–874. [CrossRef]
Sensors 2018, 18, 3531
43. 44. 45.
46.
47.
48.
49. 50. 51.
52.
53.
54.
55.
12 of 12
Bilby, D.; Kubinski, D.J.; Maricq, M. Current amplification in an electrostatic trap by soot dendrite growth and fragmentation: Application to soot sensors. J. Aerosol Sci. 2016, 98, 41–58. [CrossRef] Burtscher, H. Physical Characterization of Particulate Emissions from Diesel Engines: A Review. J. Aerosol Sci. 2005, 36, 896–932. [CrossRef] Dunne, L.J.; Sarkar, A.K.; Kroto, H.W.; Munn, J.; Kathirgamanathan, P.; Heinen, U.; Fernandez, J.; Hare, J.; Reid, D.G.; Clark, A.D. Electrical, Magnetic and Structural Characterization of Fullerene Soots. J. Phys. Condens. Matter 1996, 8, 2127–2141. [CrossRef] Sielicki, P.; Janik, H.; Guzman, A.; Namie´snik, J. Grain Type and Size of Particulate Matter from Diesel Vehicle Exhausts Analysed by Transmission Electron Microscopy. Environ. Technol. 2012, 33, 1781–1788. [CrossRef] [PubMed] Huffman, G.P.; Huggins, F.E.; Shah, N.; Huggins, R.; Linak, W.P.; Miller, C.A.; Pugmire, R.J.; Meuzelaar, H.L.; Seehra, M.S.; Manivannan, A. Characterization of Fine Particulate Matter Produced by Combustion of Residual Fuel Oil. J. Air Waste Manag. Assoc. 2000, 50, 1106–1114. [CrossRef] [PubMed] Luo, C.H.; Lee, W.M.; Liaw, J.J. Morphological and Semi-Quantitative Characteristics of Diesel Soot Agglomerates Emitted from Commercial Vehicles and a Dynamometer. J. Environ. Sci. 2009, 21, 452–457. [CrossRef] Monahan, P. Cleaning Up Diesel Pollution: Emissions from Off-Highway Engines by State; Union of Concerned Scientists: Cambridge, MA, USA, 2003. Krinke, T.J.; Deppert, K.; Magnusson, M.H.; Schmidt, F.; Fissan, H. Microscopic Aspects of the Deposition of Nanoparticles from the Gas Phase. J. Aerosol Sci. 2002, 33, 1341–1359. [CrossRef] Ritter, T.; Wiegärtner, S.; Hagen, G.; Moos, R. Simulation of a thermoelectric gas sensor that determines hydrocarbon concentrations in exhausts and the light-off temperature of catalyst materials. J. Sens. Sens. Syst. 2017, 6, 395–405. [CrossRef] Messerer, A.; Niessner, R.; Pöschl, U. Thermophoretic Deposition of Soot Aerosol Particles under Experimental Conditions Relevant for Modern Diesel Engine Exhaust Gas Systems. J. Aerosol Sci. 2003, 34, 1009–1021. [CrossRef] Lutic, D.; Pagels, J.; Bjorklund, R.; Josza, P.; Visser, J.H.; Grant, A.W.; Johansson, M.L.; Paaso, J.; Fägerman, P.-E.; Sanati, M.; et al. Detection of Soot Using a Resistivity Sensor Device Employing Thermophoretic Particle Deposition. J. Sens. 2010, 2010, 421072. [CrossRef] Hagen, G.; Spannbauer, C.; Moos, R. Electrophoretic and thermophoretic effects on conductometric soot sensing: Special challenges when using synthetic soot. In Proceedings of the 17th International Meeting on Chemical Sensors (IMCS 17), Vienna, Austria, 15–19 July 2018. [CrossRef] Fragkiadoulakis, P.; Geivanidis, S.; Samaras, Z. Modeling a resistive soot sensor by particle deposition mechanisms. J. Aerosol Sci. 2018, 123, 76–90. [CrossRef] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
