sensors Article
Development of Potentiometric Sensors for C2H4 Detection Fidel Toldra-Reig
and Jose M. Serra *
Instituto de Tecnología Química, Universitat Politècnica de València—Consejo Superior de Investigaciones Científicas, Av. Los Naranjos s/n, E-46022 Valencia, Spain;
[email protected] * Correspondence:
[email protected]; Tel.: +34-963-879-448
Received: 2 July 2018; Accepted: 29 August 2018; Published: 7 September 2018
Abstract: Gas exhaust emissions in vehicles are increasingly restrictive in EU and USA. Diesel engines are particularly affected by limitation in hydrocarbons and NOx concentrations. This work presents a screening of working electrode materials to develop a potentiometric sensor, with the most promising material to detect being C2 H4 at 550 ◦ C. The device consists of a dense 8YSZ (8 mol% Y2 O3 stabilized ZrO2 ) disk as oxide-ion conducting electrolyte, whereas platinum is screen-printed in the back face as reference electrode. As working electrode, several materials such as Fe0.7 Cr1.3 O3 , ZnCr2 O4 , Fe2 NiO4 , La0.8 Sr0.2 CrO3−δ (LSC), La0.8 Sr0.2 MnO3 (LSM), and NiO+5%wt Au were tested to detect C2 H4 . Sensor voltage was measured for several concentrations of C2 H4 and CO as these are two of the major oxidizable compounds in a diesel exhaust gas. Fe0.7 Cr1.3 O3 was selected as the most promising material because of its response to C2 H4 and CO. Not only is the response to the individual analytes important, but the C2 H4 cross-sensitivity toward CO is also important. Fe0.7 Cr1.3 O3 showed a good performance to C2 H4 , with low cross-sensitivity to CO. In addition, when 0.16 ppm of phenanthrene is added, the sensor still has a slightly better response to C2 H4 than to CO. Nevertheless, the sensor exposure to high concentrations (>85 ppm) of polycyclic aromatic hydrocarbons led to signal saturation. On the other hand, the operation in wet conditions induces lower sensor sensitivity to C2 H4 and higher cross-sensitivity toward CO increase, i.e., the sensor response becomes similar for C2 H4 and CO. Keywords: hydrocarbon; ethylene; potentiometric; sensor; YSZ; electrochemical cell
1. Introduction In the recent past years, the detection of hydrocarbons in exhaust gas of diesel cars has been increasing in importance because hydrocarbons, as well as other pollutants, are related to health issue problems in humans [1]. Then, both American and European legislators are reducing the permitted emissions limits of these pollutants [2–4]. Nowadays, there are restrictions because of the impossibility to detect HCs at lower concentrations. As a health issue, more restrictive legislations will be enforced if sensors able to detect HCs at this lower concentration level are technically and commercially available. Potentiometric hydrocarbon sensors are the most appropriate candidates because of the high temperature of the exhaust gases and the gas composition and its simplicity and low cost. This type of sensor works as a solid-state electrochemical cell, where oxygen is reduced on the reference electrode and the oxygen ions generated diffuse through the solid electrolyte, e.g., 8YSZ (8 mol% Y2 O3 -stabilized ZrO2 ) and CGO (cerium gadolinium oxides), to oxidize the target gas (analyte) on the working electrode. Both electrodes—working and reference—are typically exposed to the same atmosphere, thus it is expected to behave as a mixed potential sensor. In fact, several reactions take place in each electrode and equilibrium is achieved. The reaction with the fastest kinetics rate will
Sensors 2018, 18, 2992; doi:10.3390/s18092992
www.mdpi.com/journal/sensors
Sensors 2018, 18, 2992
2 of 14
determine the electrochemical behavior of the sensor. On the other hand, the reference electrode could be exposed to another chamber where only oxygen in air is present. Mixed potential sensors are preferred because both reference and working electrodes can be exposed to the same gas during operation. This makes the sensor suitable for exhaust automotive detection, e.g., detection of CO, NOx , HCs, NH3 [5–7]. In this sensor, the working electrode should have catalytic and sensing activity towards hydrocarbons, and low cross-sensitivity towards other gas components. On the other hand, the reference electrode should have catalytic activity toward oxygen reduction but must not convert other gas molecules (analytes). A proper selection of selective materials for the working and reference electrode is essential. In the present work, several materials were screened to reach materials suitable for C2 H4 detection with limited cross-sensitivity toward CO. The sensor device consists of a disk-shaped working solid 8YSZ membrane electrolyte [8–10] with a reference and a working electrode deposited on each face of the disk. Both electrodes are highly porous and were deposited by screen-printing. In literature, materials such as LaCoO3 [11], Nb2 O5 [12], Au/8YSZ [8], SnO2 [13], Cr2 O3 [13], La2 CuO4 [14], TiO2 (+Pd) [15], La1−x Srx Cr1−y Gay O3−δ [16], Pt/YSZ, MgAl2 O4 [17], and NiO/Au [18]. Thus, as working electrodes, several perovskites and metal oxides are studied in this work, similar to those aforementioned. The cell voltage is measured in atmospheres containing C2 H4 and CO, as they are the most abundant partially oxidized compounds in diesel exhaust gas. A 6% fixed O2 concentration is employed, while the influence of water and aromatics is also studied. The most promising material is selected for a full characterization for ethylene detection purposes. One novel approach of the present work is the study of the cross-sensitivity: The sensor is exposed to a fixed concentration of one analyte while concentration pulses of a second analyte are performed. In literature, the cross-sensitivity is usually evaluated comparing the sensitivity of each element individually. Taking into account that competitive reactions can take place simultaneously in each electrode, single-gas measurements could not be adequate to evaluate the cross-sensitivity. Porous platinum is usually employed as reference electrode due to its high catalytic activity, thermal stability, and common use in sensor electrodes [19]. 2. Materials and Methods 2.1. Synthesis of Electrode Materials The starting precursors for the synthesis of working electrode materials were commercial nitrates from Sigma Aldrich. A sol-gel chemical route was followed to obtain single-phase perovskites and spinels. First, stoichiometric quantities of nitrate precursors were fully dissolved in water. After that, citric acid (Sigma Aldrich, St. Louis, MO, USA) was added as a chelating agent to prevent partial segregation or precipitation of the metal components, and then ethylene glycol was added to promote the polymerization of the chelating agent and produce an organometallic polymer resin (in a molar ratio 1:2:4 with respect to nitrates solution, citric acid, and ethylene glycol, respectively). This complexation step is followed by dehydration at low temperature (up to 200 ◦ C) and, finally, thermal decomposition in air of the precursors at 600 ◦ C led to the formation of nanosized crystalline phases. The obtained powders were sintered at 1350 ◦ C for 10 h and milled to produce the target crystalline phase and a homogenous particle size. Then, the material was mixed with an organic binder (terpineol with a 6%wt ethylcellulose) in a 1:2 weight ratio and passed through a three-roller mill to produce screen-printing inks. As reference electrode, a commercial platinum powder (particle size 0.2 to 1.8 µm, from Mateck (Jülich, Germany)) is employed and the same procedure is followed to obtain the ink. Several materials, such as Fe0.7 Cr1.3 O3 , ZnCr2 O4 , Fe2 NiO4 , La0.8 Sr0.2 CrO3−δ (LSC), La0.8 Sr0.2 MnO3−δ (LSM), and NiO+5%wt Au, are employed as working electrodes. The aim was to find a material that provides a good response to ethylene. Once the most promising material was selected, further analyses were carried out with this configuration. Only in the case of the Fe0.7 Cr1.3 O3 , as the material alone did not attach properly to the 8YSZ electrolyte, was the powder was mixed with 8YSZ powder (Tosoh, Tokyo, Japan) to ensure a good
Sensors Sensors 2018, 2018, 18, 18, 2992 x FOR PEER REVIEW
33of of14 14
electrode attachment. For this purpose, both powders in a 1:1 volume ratio were ball-milled for 24 h. electrode attachment. For this purpose, both powders in a 1:1 volume ratio were ball-milled for 24 h. Furthermore, this leads to an increase of the triple phase boundary length, i.e., the reaction sites for Furthermore, this leads to an increase of the triple phase boundary length, i.e., the reaction sites for the the electrochemical hydrocarbon oxidation. electrochemical hydrocarbon oxidation. 2.2. Fabrication Fabrication of of the the Sensor Sensor Device Device 2.2. Individual 8YSZ 8YSZdisks disks 20 mm of diameter were employed as electrolyte for the sensor Individual of of 20 mm of diameter were employed as electrolyte for the sensor assembly. assembly. The 8YSZ powder was uniaxially at 50disks kN into then at sintered atfor 1350 The 8YSZ powder was uniaxially pressed atpressed 50 kN into anddisks then and sintered 1350 ◦ C 10°C h. for 10 h. The device (Figure 1) is completed by screen-printing both circular electrodes (9 mm in The device (Figure 1) is completed by screen-printing both circular electrodes (9 mm in diameter). diameter). First, the workingiselectrode is printed and 1150 °C,platinum and then reference platinum electrode reference First, the working electrode printed and sintered atsintered 1150 ◦ C,atand then electrode in face the back face at the same conditions. porous gold layerpattern) (mesh pattern) is printedisinprinted the back at the same conditions. Finally, Finally, a porousa gold layer (mesh acting acting as current collector is screen-printed on top of the sintered working electrode, and as current collector is screen-printed on top of the sintered working electrode, and sinteredsintered at 900 ◦at C 900 °C for 2 h. Gold is not expected to affect the sensor response as its catalytic activity depends for 2 h. Gold is not expected to affect the sensor response as its catalytic activity depends critically on criticallysize. on particle size. gold isinert chemically inert known and is well be acatalyst. very poor catalyst. particle Bulk gold is Bulk chemically and is well to beknown a veryto poor It only has It only has activity when employed as nanoparticle and at low temperatures [20,21]. Silver ink to is activity when employed as nanoparticle and at low temperatures [20,21]. Silver ink is employed employed to ensure contact between gold lead wires. ensure contact between electrode andelectrode gold leadand wires.
C2H4 CO2 + H2O CO
CO2
WE 8YSZ
O2O2-
RE O2 Figure 1. Scheme of the sensor. The device consists of a dense 8YSZ electrolyte with platinum reference Figure 1.(bottom) Scheme and of the sensor. The device consists of a dense electrode distinct materials as working electrode (top).8YSZ electrolyte with platinum reference electrode (bottom) and distinct materials as working electrode (top).
2.3. Characterization of Materials and Sensor Device 2.3. Characterization of Materials and Sensor Device A PANalytical Cubix fast diffractometer, using CuKα1 radiation (λ = 1.5406 Å) and an X0 Celerator A PANalytical Cubix fast diffractometer, using α1 radiation (λof = 1.5406 Å) and an X′Celerator detector in Bragg–Brentano geometry was used for CuK the identification the crystalline phases. XRD detector recorded in Bragg–Brentano geometry used theanalyzed identification of the Highscore crystalline Plus phases. XRD patterns in the 2θ ◦ range from was 10◦ to 90◦ for were using X’Pert software patterns recorded in Malvern, the 2θ° range from and 10° energy-dispersive to 90° were analyzed X’Pert Highscore Plus (Malvern Panalytical, UK). SEM X-rayusing spectroscopy (EDS) using a software (Malvern Malvern, UK). SEM and energy-dispersive X-rayGermany) spectroscopy ZEISS Ultra55 field Panalytical, emission scanning electron microscope (ZEISS, Oberkochen, was(EDS) used using a ZEISS Ultra55 field emission scanning to analyze fracture cross-sections of the sinteredelectron materialmicroscope before and(ZEISS, after theOberkochen, permeation Germany) test. SEM was used to analyze fracture cross-sections of thetosintered before and after thecontrast permeation backscattered electrons detector (BSD) was used providematerial images with compositional that test. SEM backscattered electrons detector (BSD) was used to provide images with compositional differentiate grains and element distribution. contrast differentiate grains and distribution. The that voltage was determined byelement measuring the potential difference between the working and The voltage was determined by measuring potential difference between thewere working reference electrode when a zero-ampere current isthe applied. Gold wires and silver ink usedand for ◦ C and reference electrode when zero-ampere current is applied. andseveral silver ink were used for contacting electrodes. Thea measurements were carried out Gold at 550wires concentrations of contacting electrodes. measurements were carried out atbetween 550 °C and concentrations of C CO wereThe measured (Figure S1). The cell voltage bothseveral electrodes was detected 2 H4 and/or C2H and/or CO Keithley were measured (FigureInstrument, S1). The cell voltage between both electrodes detected by a 4multimeter, 3706 (Keithley Cleveland, OH, USA). The responsewas of the sensor bycell a , multimeter, Keithley (V mV) was defined as: 3706 (Keithley Instrument, Cleveland, OH, USA). The response of the sensor (Vcell, mV) was defined as: Vcell = Va − Vb (1)
Vcell = Va − V where Va and Vb are the voltage of the sensor exposed tob (a) analytes with the background gas of(1) 6% O (in Ar balance), and (b) background gas alone (6% O in Ar balance), respectively. 2 2 analytes with the background gas of 6% O2 where Va and Vb are the voltage of the sensor exposed to (a) Mass flow controllers were used to obtain different gas mixtures. The total gas flow rate was (in Ar balance), and (b) background gas alone (6% O2 in Ar balance), respectively. 550 mL/min with a 6% of oxygen, varying C H and CO concentrations and Ar as balance. Oncerate the 2 4 Mass flow controllers were used to obtain different gas mixtures. The total gas flow was 550 mL/min with a 6% of oxygen, varying C2H4 and CO concentrations and Ar as balance.
Sensors 2018, 18, 2992
4 of 14
sample is stabilized at 550 ◦ C, C2 H4 or CO concentration was varied from 50 ppm (used as base gas) to 100, 150, and 200 ppm. The same test was performed for each gas with 200 ppm of the other gas as background in order to check cross-sensitivity. The same tests are performed with a background of water (3%) and polyaromatics: Toluene (28,947 ppm), methylnaphtalene (88 ppm), and phenanthrene (0.16 ppm). The gas flow is saturated in these elements at room temperature and then the selection of polyaromatics is limited to (1) their vapor pressure, and (2) the fact that polyaromatics concentration in exhaust gases is usually in the range from 0 to 10 ppm [22–24]. The aromatics are selected according to its vapor pressure to provide a concentration close to the desired range: One compound with a much higher concentration than expected in an exhaust gas (toluene), as it could be interesting to check the performance under extreme conditions, and two compounds closer to the range in exhaust gases (methylnaphthalene and phenanthrene). Electrochemical impedance spectroscopy analysis was carried for each analyte at 200 ppm using an Autolab PGSTAT204 (Metrohm Autolab, Utrecht, The Netherlands) with a FRA32M module. The frequency was changed from 0.03 Hz to 1 MHz. Prior to the EIS measurement, open circuit voltage (OCV) is measured and this bias voltage is applied for the EIS measurement. The polarization resistance is measured as the difference between the high frequency and low frequency intercepts with the real axis of the impedance (Z’). Coke formation on the surface of the electrode was studied by Raman spectroscopy. Raman spectrometer is equipped with a Leica DMLM microscope (Leica Microsystems, Wetzlar, Germany) and a 785-nm Ar+ ion laser as an excitation source. A 50× objective of 8-mm optical length was used to focus the depolarized laser beam on a spot of 3 µm in diameter. A charged coupled device (CCD) array detector was used for the Raman scattering collection. 3. Results and Discussion 3.1. Microstructural Characterization Analysis of the X-ray diffraction (XRD) patterns (Figure 2) of the different as-sintered electrode powders confirms that the target crystalline phase was obtained for each material. Diffraction peaks Sensors 2018, 18, x FOR PEER REVIEW 5 of 14 corresponding to secondary phases are not observed.
Figure 2. X-ray diffraction patterns of the materials studied at room temperature once they Figure 2. X-ray diffraction patterns of the materials studied at room temperature once they are synthesized. are synthesized.
SEM analysis (Figure 3 and Figure S2 in Supporting Information) shows the structural and morphologic characteristics of the built sensor device. The fracture cross-section of the porous working electrode (Figure 3a,b) reveals a distinct grain size distribution for Fe0.7 Cr1.3 O3 and 8YSZ,
Sensors 2018, 18, 2992
5 of 14
i.e., Fe0.7 Cr1.3 O3 grains are bigger. Nevertheless, a homogeneous distribution of grains is reached throughout the 10 µm-thick electrode. No reaction interfaces are detected between electrodes and electrolyte. Neither reaction between the phases nor impurities on the grain boundaries are detected (Figure 3b–d). The reference electrode is a 7 µm-platinum layer made of highly-coarsened grains with good adhesion to the electrolyte. Figure 3c shows a view of the complete device with a ~1mm-thick dense 8YSZ electrolyte. For all electrodes with different compositions, a proper attachment of electrodes 2. X-ray diffraction of the materials at room once they toFigure the 8YSZ substrate was patterns achieved, excepting Lastudied (LSC) electrode that are wassynthesized. detached. 0.8 Sr0.2 CrO 3−δ temperature
Figure as WE WE and and Pt Pt as as RE. RE. Figure 3. 3. SEM SEM image image of of the the device device cross-section cross-sectioncorresponding correspondingtotoFeCr FeCr22O O44/8YSZ /8YSZ as (a) WE electrode cross-section with FeCr O and 8YSZ grains. (b) Interface WE-electrolyte where (a) WE electrode cross-section with FeCr22O44 and 8YSZ grains. (b) Interface WE-electrolyte where aa good good attachment attachment of of the the layer layer can can be beobserved. observed. Gold Gold collector collector layer layer at atthe thetop topas aswell wellas asFeCr FeCr22O O44 and and 8YSZ be be observed. (c) Complete device Fe0.7 Cr /8YSZ || 8YSZ || Pt can 1.3 O23O 8YSZ grain graindistribution distributioncan can observed. (c) Complete device FeCr 4/8YSZ//8YSZ//Pt can be be observed. Sintered platinum reference observed. observed. (d)(d) Sintered platinum reference cancan be be observed.
3.2. Electrochemical Characterization 3.2. Electrochemical Characterization In the presence of CO, hydrocarbons, other reducing agents, and oxygen, several oxidation and In the presence of CO, hydrocarbons, other reducing agents, and oxygen, several oxidation and reduction reactions take place at the interface among electrode, electrolyte, and gas phase (mixed reduction reactions take place at the interface among electrode, electrolyte, and gas phase (mixed potential sensor). In a potentiometric sensor, a zero-current is applied. Thus, competitive reactions potential sensor). In a potentiometric sensor, a zero-current is applied. Thus, competitive reactions take place in both electrode and therefore, kinetics will be a key point. When a steady-state between take place in both electrode and therefore, kinetics will be a key point. When a steady-state between the anodic reaction (Equation (3) or Equation (4)) and the cathodic reaction Equation (2) is achieved, the anodic reaction (Equation (3) or Equation (4)) and the cathodic reaction Equation (2) is achieved, a mixed-potential is established. The difference between the mixed-potential established in both a mixed-potential is established. The difference between the mixed-potential established in both electrodes will give rise to the voltage of the cell [25]. Then, the sensor performance can be enhanced electrodes will give rise to the voltage of the cell [25]. Then, the sensor performance can be enhanced with a proper selection of the materials: Selective material to the desired target analyte in the WE, and with a proper selection of the materials: Selective material to the desired target analyte in the WE, a material active to oxygen as RE (Figure 1). Note that both electrodes are exposed to the same gas and a material active to oxygen as RE (Figure 1). Note that both electrodes are exposed to the same mixture, i.e., RE is not exposed to a reference gas (air). In addition, heterogeneous catalytic conversion gas mixture, i.e., RE is not exposed to a reference gas (air). In addition, heterogeneous catalytic in the WE of the target gases (analytes) with locally adsorbed O2 can occur at the same time Equations conversion in the WE of the target gases (analytes) with locally adsorbed O 2 can occur at the (5) and (6). This reaction will compete with the (sensing) electrochemical reactions and affect the sensor same time Equations (5) and (6). This reaction will compete with the (sensing) electrochemical response if the electrode is not active towards electrochemical oxidation of C2 H4 . These are the specific reactions and affect the sensor response if the electrode is not active towards electrochemical reactions taken into account [26]. oxidation of C2H4. These are the specific reactions taken into account [26]. 1 O2 + 2e− → O2− (2) 2 CO + O2− → CO2 + 2e−
(3)
C2 H4 + 6O2− → 2CO2 + 2H2 O + 12e−
(4)
C2 H4 + 3O2 → 2CO2 + 2H2 O
(5)
CO + 0.5O2 → CO2
(6)
hydrocarbon sensing applications in an environment such as diesel exhaust. Thus, a high response to C2H4 and a low response to CO are required. Sensors were exposed to several concentrations of each gas. In a first screening, the sensors were exposed to a concentration range from 200 to 1000 ppm and, in a second step, the sensors were exposed to a concentration range from 50 to 200 ppm. As the sensor responses are linear with the analyte concentration within the studied range, the Sensors 2018, 18, 2992 6 of 14 sensor sensitivity (Figure 4) is defined as the slope dVcell/dCanalyte (mV/ppm) and calculated as a linear regression. NiO+5%wt Au, FeNiO4, and NiO show higher response to CO rather than C2H4, and were In ainvestigated. first set of experiments, several perovskites, spinels, other oxides were studied further LSC and LSM show low response withand low sensitivity toward C2H4employing . ZnCr2O4 platinum as reference electrode. The aim is to identify promising materials for hydrocarbon sensing and Fe0.7Cr1.3O3 exhibit promising C2H4 sensitivity, i.e., high response to C2H4 and low response to applications in an ZnCr environment suchisasnot diesel exhaust. Thus, a high response tohighest C2 H4 and a low CO. Nevertheless, 2O4 response stable with time despite providing the response response to CO are required. Sensors were exposed toand several concentrations of each gas. In a first to C2H4. Thus, the sensor response is not reproducible not appropriate for long-term operation. screening, the sensors were exposed to a concentration range from 200 to 1000 ppm and, in a second This could be due to an ageing effect, e.g., sintering and coking (Figure S3 in Supporting Information). step, the exposed a concentration range from 50 tobecause 200 ppm. Thus, Fe0.7sensors Cr1.3O3 were is a priori the to most promising sensing material of the stable response and As the sensor responses appropriate response to C2H4.are linear with the analyte concentration within the studied range, the sensor is3 with defined as ensures the slopegood dVcell /dCanalyteto (mV/ppm) and calculated Thesensitivity mixture of(Figure Fe0.7Cr4) 1.3O 8YSZ attachment the electrolyte and enablesastoa linear regression. NiO+5%wt Au, available FeNiO4 , and NiOelectrochemical show higher response to COoxidation, rather than C2the H4 , enlarge the number of active sites for the hydrocarbon i.e., and were further investigated. LSC and LSM show low response with low sensitivity toward C H4 . triple phase boundary (TPB) length. Therefore, apart from the intrinsic catalytic activity of 2the ZnCr2 O4 material, and Fe0.7the Cr1.3 O3 exhibit promising i.e.,ofhigh response to low electrode sensor performance canCbe because the increment ofCthe 2 Hboosted 4 sensitivity, 2 HTPB 4 andarea. response to CO. response is not stable with time despite providing the highest Then, in order toNevertheless, disclose if theZnCr good2 Operformance of Fe 0.7 Cr 1.3 O 3 is due to the catalytic activity of the 4 response to C H . Thus, the sensor response is not reproducible and not appropriate for long-term material itself 2or 4because of the increment of the TPB area, another tested material (LSC) is mixed operation. Thisvolume could beratio) due to anexposed ageing effect, e.g., sintering and coking S3 in Supporting with 8YSZ (1:1 and to C2H 4. When compared to the(Figure bare LSC electrode, no Information). Thus, Fe0.7 Cr1.3(Figure O3 is a S4) priori most promising material because of1.3the stable significant effect is observed andthe therefore the goodsensing performance of the Fe0.7Cr O3-8YSZ response and appropriate response H34activity. . can be attributed to the intrinsic Fe0.7to CrC1.32O
Figure 4. Sensor sensitivity (mV/ppm) to C2 H4 (red) and CO (black) individually for several materials used as4.working electrode. (mV/ppm) The sensor to sensitivity is and defined Vcell /analyte concentration Figure Sensor sensitivity C2H4 (red) CO as (black) individually for several(mV/ppm) materials in the linear range of sensor response and calculated as a linear regression. ZnCr Fe2 NiO4 2 O4 and used as working electrode. The sensor sensitivity is defined as Vcell/analyte concentration (mV/ppm) are evaluated in the range 200–1000 ppm, while the rest of the materials (Fe0.7 Cr1.3 O3 , LSC, LSM, and NiO+5%wt Au) are evaluated in the range 50–200 ppm.
The mixture of Fe0.7 Cr1.3 O3 with 8YSZ ensures good attachment to the electrolyte and enables to enlarge the number of active sites available for the electrochemical hydrocarbon oxidation, i.e., the triple phase boundary (TPB) length. Therefore, apart from the intrinsic catalytic activity of the electrode material, the sensor performance can be boosted because of the increment of the TPB area. Then, in order to disclose if the good performance of Fe0.7 Cr1.3 O3 is due to the catalytic activity of the material itself or because of the increment of the TPB area, another tested material (LSC) is mixed with 8YSZ (1:1 volume ratio) and exposed to C2 H4 . When compared to the bare LSC electrode, no significant effect is observed (Figure S4) and therefore the good performance of the Fe0.7 Cr1.3 O3 -8YSZ can be attributed to the intrinsic Fe0.7 Cr1.3 O3 activity. Sensors based on 8YSZ-Fe0.7 Cr1.3 O3 working electrode were further characterized by exposing them to C2 H4 and CO mixtures with several concentrations. Specifically, a background concentration (200 ppm) of one analyte is set when varying the concentration of the other analyte. Figure 5 shows the recorded Vcell as a function of time for series of stepwise changes in gas composition, and illustrates
evaluated in the range 200–1000 ppm, while the rest of the materials (Fe0.7Cr1.3O3, LSC, LSM, and NiO+5%wt Au) are evaluated in the range 50–200 ppm.
Sensors based on 8YSZ-Fe0.7Cr1.3O3 working electrode were further characterized by exposing them to C2H4 and CO mixtures with several concentrations. Specifically, a background concentration Sensors 2018, 18, 2992 7 of 14 (200 ppm) of one analyte is set when varying the concentration of the other analyte. Figure 5 shows the recorded Vcell as a function of time for series of stepwise changes in gas composition, and illustrates high response sensitivity to Cto 2H4 (Figure 5a), together with low cross-sensitivity toward CO (Figure 5b). high response sensitivity C2 H4 (Figure 5a), together with low cross-sensitivity toward CO (Figure 5b). Indeed, the sensor hardly responds to to changes changes in in CO CO concentration concentration when when C C2H 4 (at 200 ppm level) is Indeed, the sensor hardly responds 2 H4 (at 200 ppm level) is present in in the the gas. gas. Further, Further, this this figure upon present figure reveals reveals that that the the reproducibility reproducibility and and stability stabilityof of the the signal signal upon equilibration, while response kinetics is rather fast. Additionally, the sensor baseline is close to zero equilibration, while response kinetics is rather fast. Additionally, the sensor baseline is close to zero and stable stable without without any any drift drift observed observed (Figure (Figure S5). S5). and
Figure 5. Response curves for changes in concentration of: (a) C2 H4 , (b) C2 H4 with 200 ppm CO Figure 5. Response curves for changes in concentration of: (a) C2H4, (b) C2H4 with 200 ppm CO background, (c) CO, and (d) CO with 200 ppm C2 H4 background. Each change in concentration background, (c) CO, and (d) CO with 200 ppm C2H4 background. Each change in concentration has has 20 min duration and goes from 50 ppm to 100, 150, and 200 ppm. The sensor consists of 20 min duration and goes from 50 ppm to 100, 150, and 200 ppm. The sensor consists of Fe0.7 Cr1.3 O3 /8YSZ as WE and Pt as RE, at 550 ◦ C with 6% of O2 . Fe0.7Cr1.3O3/8YSZ as WE and Pt as RE, at 550 °C with 6% of O2.
Figure 6 plots the equilibrium Vcell as a function of CO/C2 H4 concentration. In dry conditions 6 plots the equilibrium cell as a function of CO/C2H4 concentration. In dry conditions (openFigure symbols in Figure 6), the sensorVresponse (Vcell ) is proportional to C2 H4 concentration regardless (open symbols in Figure 6), the sensor response (Vcell ) ischanges proportional to C2negligible. H4 concentration regardless of the CO background, whereas the response to CO is nearly The same study of the CO background, whereas the response to CO changes is nearly negligible. The same was was repeated by adding 3% of water (saturation of inlet gases at room temperature). Thestudy addition repeated adding 3% ofin water (saturation of inlet gases at room temperature). addition water of water by (solid symbols Figure 6) strongly increases the sensitivity of CO The while slightlyofaffects (solid symbols in Figure 6) strongly increases the sensitivity of CO while slightly affects sensitivity to sensitivity to C2 H4 and, consequently, leads to higher cross-sensitivity to CO (Table 1). Two effects C 2H4 and, consequently, leads to higher cross-sensitivity to CO (Table 1). Two effects may be may be responsible for this behavior: (1) Water could promote CO conversion by removing surface responsible for thisorbehavior: couldand promote CO conversion removing coke from coke from surface avoiding(1) itsWater formation; (2) water enables the by formation of surface H2 on the surface surface or avoiding its formation; and (2) water enables the formation of H 2 on the surface of both of both WE and RE (Pt) via water gas-shift reaction (WGS, Equation (7)). The electrochemical oxidation WEthe and RE (Pt) viaHwater gas-shift reaction (WGS, Equation (7)). The electrochemical oxidation of of locally-built 2 on the electrodes is kinetically more favored than CO oxidation and this may the locally-built H2 on the electrodes is kinetically more favored than CO oxidation and this may have have a notable impact in the sensor response to CO. a notable impact in the sensor response to CO. HH CO2+ +HH2 2 OO++CO CO → → CO
(7) (7)
Sensors 2018, 18, x FOR PEER REVIEW 2992
of 14 14 88 of
Figure 6. 6. EMF EMFSensor Sensorresponse responseasasa function a function analyte concentration (a) C and H4 (b) andCO. (b) The CO. Figure of of analyte concentration for:for: (a) C 2H 4 2 The effect of a background concentration (200of ppm) of the analyte second and analyte and 3% is ofalso water is also effect of a background concentration (200 ppm) the second 3% of water depicted. depicted. Each point is the main of the respective 20 min change in response. Each point is the main of the respective 20 min change in response. Table 1. Sensitivity of the sensor to C2 H4 , carbon monoxide, and each analyte with 200 ppm of Table 1. Sensitivity of the sensor C2dry H4, carbon monoxide, and eachasanalyte 200 ppm background of the second analytetofor and wet (3%) conditions, well aswith for 0.16 ppm of of background of the second analyte for dry and wet (3%) conditions, as well as for 0.16 ppm of C14Has 10. C14 H10 . Sensor sensitivity is defined as Vcell /analyte concentration (mV/ppm) and calculated Sensor is defined as Vcell/analyte concentration (mV/ppm) and calculated as a linear regression. a linearsensitivity regression.
C14H10 Sensitivity Dry(mV/ppm) Sensitivity WetWet Sensitivity C14 H10 Sensitivity Analyte Analyte Dry Sensitivity Sensitivity (mV/ppm) (mV/ppm) (mV/ppm) (mV/ppm) (mV/ppm) C2H4
C2 H4 CO CO C2 H4 with 200 ppm of CO C2H4 with 200 ppm of CO CO with 200 ppm of C2 H4 CO with 200 ppm of C2H4
−1 1.2 × 101.2 × 10−1 2.4 × 10−2 −2 2.4 × 10 8.4 × 10−2 8.4 × 10−2 6.6 × 10−3 6.6 × 10−3
×− 102−2 9.4 9.4 × 10 −2−2 8.8 8.8 × 10 × 10 5.1 × 10−2−2 5.1 × 10 5.8 × 10−2 5.8 × 10−2
2 × 10−2 9.0 × 10−9.0 2 7.1 × 10−7.1 × 10−2 6.1 × 10−2 6.1 × 10−2 5.1 × 10−2 5.1 × 10−2
Figure 7 displays the electrochemical impedance spectra (Nyquist and Bode plots, frequency Figure 7 displays the electrochemical impedance spectra (Nyquist and Bode plots, frequency range 0.03 Hz to 1 MHz) of the sensor exposed to 200 ppm of C2 H4 or CO, both in dry and wet range 0.03 Hz to 1 MHz) of the sensor exposed to 200 ppm of C2H4 or CO, both in dry and wet conditions. Electrochemical impedance spectroscopy (EIS) analysis provides information related with conditions. Electrochemical impedance spectroscopy (EIS) analysis provides information related the processes taking place in the device (ionic transport, solid-solid interfacial resistance, gas-solid with the processes taking place in the device (ionic transport, solid-solid interfacial resistance, gassurface reactions) since they may exhibit distinct characteristic frequencies. For the present case, the solid surface reactions) since they may exhibit distinct characteristic frequencies. For the present case, variation of the analyte concentration or type may induce changes in the impedance contributions the variation of the analyte concentration or type may induce changes in the impedance contributions only related to gas-solid surface reactions and appearing at the lowest frequencies. The larger the only related to gas-solid surface reactions and appearing at the lowest frequencies. The larger the impedance, the slower (or less favored) the specific surface reaction (oxidation of analytes in this impedance, the slower (or less favored) the specific surface reaction (oxidation of analytes in this case) is. Although the results are rather similar for both analytes, there is a slight difference at low case) is. Although the results are rather similar for both analytes, there is a slight difference at low frequencies for both analytes (impedance is slight lower in C H ) that can be linked to the results frequencies for both analytes (impedance is slight lower in C22H44) that can be linked to the results observed in the potentiometric characterization. In dry conditions (Figure 7a,b), the shape of the arc at observed in the potentiometric characterization. In dry conditions (Figure 7a,b), the shape of the arc high frequencies is similar for both analytes, therefore the ionic transport remains almost unaffected. at high frequencies is similar for both analytes, therefore the ionic transport remains almost On the other hand, the response impedance (Z’) at low frequencies is sensitive to changes in the unaffected. On the other hand, the response impedance (Z’) at low frequencies is sensitive to changes gas composition and this suggests that the sensor response depends on electrochemical reactions at in the gas composition and this suggests that the sensor response depends on electrochemical the surface of Fe0.7 Cr1.3 O3 /8YSZ and interface with 8YSZ electrolyte [27–31]. Specifically, the low reactions at the surface of Fe0.7Cr1.3O3/8YSZ and interface with 8YSZ electrolyte [27–31]. Specifically, frequency response is higher in CO than in C2 H4 (Table 2) and is in line with the observed sensor the low frequency response is higher in CO than in C2H4 (Table 2) and is in line with the observed behavior in dry conditions. In wet conditions (Figure 7c,d), the impedance arc at low frequency is sensor behavior in dry conditions. In wet conditions (Figure 7c,d), the impedance arc at low practically unaltered in C2 H4 but it is reduced in CO (Table 2). Water has an effect on the electrochemical frequency is practically unaltered in C2H4 but it is reduced in CO (Table 2). Water has an effect on the reactions taking place in both the electrode and electrode-electrolyte interface, as observed previously electrochemical reactions taking place in both the electrode and electrode-electrolyte interface, as in the potentiometric characterization. Thus, the electrochemical CO evolution is ameliorated on the observed previously in the potentiometric characterization. Thus, the electrochemical CO evolution electrode surface thanks to the presence of water, likely mediated by locally formed H2 (Equation (7)). is ameliorated on the electrode surface thanks to the presence of water, likely mediated by locally Therefore, the decrease in polarization resistance (Rp) results in higher sensitivity to CO and therefore formed H2 (Equation (7)). Therefore, the decrease in polarization resistance (Rp) results in higher in loss of the desired sensor performance to C2 H4 . sensitivity to CO and therefore in loss of the desired sensor performance to C2H4.
Sensors Sensors2018, 2018,18, 18,2992 x FOR PEER REVIEW
9 of 1414 9 of
Figure 7. Impedance spectrometry study in dry and wet conditions for Ar+6%O2 , 200 ppm C2 H4 , and Figure Impedance spectrometry study dryin and conditions for(c) Ar+6%O 2, 200 2H 4, and 200 ppm7.CO. (a) Nyquist plot and (b) Bodeinplot drywet conditions, and Nyquist plotppm and C (d) Bode 200 ppm CO. (a) Nyquist plot and (b) Bode plot in dry conditions, and (c) Nyquist plot and (d) Bode plot in wet conditions with a 3% of H2 O. plot in wet conditions with a 3% of H2O. Table 2. Polarization resistance (Rp) as a function of analyte (C2 H4 and carbon monoxide) in dry and wet conditions. In bothresistance conditions(Rp) the as concentration the analytes ppm.monoxide) in dry and and200 carbon Table 2. Polarization a function ofofanalyte (C2H4 are
wet conditions. In both conditionsAnalyte the concentration are 200 ppm. (200 ppm) of the analytes Rp (ohm) dry wet
dry
C2 HAnalyte 4 CO (200 ppm)
Rp 5112.52 5548.20 (ohm)
C2 H4C2H4 CO CO
5079.02 5112.52 5257.03 5548.20
C2H4
5079.02
As a final test of the sensorwet based on 8YSZ-Fe0.7 Cr1.3 O3 working electrode, the effect of the CO 5257.03 exposure to minor amounts of different aromatic hydrocarbons on the potentiometric response was studied in dry (Table 3), ason they are potential in exhaustthe gases thatofcan As a final testconditions of the sensor based 8YSZ-Fe 0.7Cr1.3O3candidates working electrode, effect the interfere sensor response to C2 H4 .aromatic For thishydrocarbons purpose, the gas feed was separately saturated exposureintothe minor amounts of different on the potentiometric response was atstudied room temperature with(Table toluene ppm), methylnaphtalene (88 ppm), andthat phenanthrene in dry conditions 3), (28,947 as they are potential candidates in exhaust gases can interfere (C H10 -0.16 ppm) at room temperature. sensor highwas sensitivity to C the in14the sensor response to C 2H4. For this The purpose, theshows gas feed separately saturated at room 14 H10 while response is still slightly C2 Hmethylnaphtalene However, the difference in10-0.16 response temperature with toluenehigher (28,947in ppm), (881). ppm), and phenanthrene (C14H ppm) 4 than in CO (Table becomes when one evaluated in the presence of10200 ppm the second analyte (Figure 8 at room small temperature. Theanalyte sensor is shows high sensitivity to C14H while theofresponse is still slightly higher and 1), i.e., the(Table sensor1).response to the C2 H is practically lost in presence of C H . Similarly, in CTable 2H4 than in CO However, difference in response becomes small when one analyte 4 14 10 toluene addition leads to high of cross-sensitivity and C(Figure toluene signal is is evaluated in the presence 200 ppm of thetoward secondCO analyte 8 and Tableresponse 1), i.e., the sensor 2 H4 , while superimposed both CO and lost C2 Hin responses, although it should be taken into consideration that response to Cover 2H4 is practically presence of C 14 H 10 . Similarly, toluene addition leads to high 4 toluene and methylnaphtalene concentration out ofresponse the rangesignal of exhaust gases. Additionally, cross-sensitivity toward CO and C2H4, while is toluene is superimposed over boththe CO sensitivity sensor toward poly-aromatics increase the number ofthat rings. This could and C2Hof 4 the responses, although it shouldseems be to taken intowith consideration toluene and be related to the higher reactivity of methylnaphtalene concentration is larger out of aromatics. the range of exhaust gases. Additionally, the sensitivity of the sensor toward poly-aromatics seems to increase with the number of rings. This could be related to the higher reactivity of larger aromatics.
Sensors 2018, 18, 2992
10 of 14
Table 3. Sensor response to several conditions: Ar+6%O2 base gas, toluene, methylnaphtalene, phenantrene, C2 H4 , and carbon monoxide. Concentration, cell voltage, and sensor sensitivity as mV/ppm are provided. Compound
Cell Voltage (mV)
Ar+6%O2 0.03 C7 H8 119.16 C11 H10 −158.75 C14 H10 0.23 Sensors 2018,C18, 2 Hx4 FOR PEER REVIEW 23.49 CO 4.82
Concentration (ppm)
Sensitivity (mV/ppm)
28,947.37 88.16 0.16 200 200
0.004 −1.80 1.44 0.12 0.02
10 of 14
Figure 8. Sensor response when 0.16 ppm of phenantrene is added as a function of analyte Figure 8. Sensor when 0.16 ppm of phenantrene is added as a function of analyte concentration, C2 Hresponse 4 in red, and CO in black. The effect of background concentration (200 ppm) concentration, C 2H4 in red, and CO in black. The effect of background concentration (200 ppm) of the of the second analyte is depicted. Each point is the average of the respective 20 min change in response. second analyte is depicted. Each point is the average of the respective 20 min change in response.
Regarding sensor aging upon test in aromatics, a single exposure to one of these aromatics gives 2 base gas, toluene, methylnaphtalene, 3. Sensor response to newly severalmeasured conditions: Ar+6%O rise toTable changes in response when without aromatics, specifically the sensitivity decays. phenantrene,step C2His 4, and monoxide. Concentration, cell air), voltage, sensor sensitivityrecover as If a regeneration donecarbon (sensor calcination at 600 ◦ C with C2 Hand and CO response 4 mV/ppm are provided. the original behavior. This suggests that coke may be formed on the surface of the electrodes. The iron spinel oxide provides Lewis acidicCell sites that have a catalytic activity toward aromatics oxidation and Voltage Concentration Sensitivity Compound oligomerization [32]. Iron oxide is more affected by aging and coke formation leading to a reduction (mV) (ppm) (mV/ppm) of its catalytic activity [33]. Thus, coke effect should be removed and probably this could be due to Ar+6%O2 0.03 a surface modification when aromatics are partly oxidized during sensor operation. The sensor was C7H8 119.16 28,947.37 0.004 exposed to working conditions at 550 ◦ C for 24 days and negligible ageing effect was noticed on the C11H10 −158.75 88.16 −1.80 sensor response. C14H10 0.23 0.16 1.44 Thus, as coke formation is suspected to happen after exposure to aromatics, Raman spectroscopy 23.49 200 0.12 C2H4 analyses were performed over the Fe0.7 Cr1.3 O3 /8YSZ electrode layer with a gold collector before and CO 4.82 200 0.02 after the test in order to check the coke formation. Raman spectroscopy is a technique that enables sensor aging presence upon test of in aromatics, a single exposure to onecarbon of theseforms. aromatics gives one toRegarding identify the potential surface species related to distinct Figure 9 rise to changes in response when newly measured without aromatics, specifically the sensitivity depicts both Raman spectra where the peaks corresponding to the spinel phase can be noticed at band 1 [34–36]. decays.ofIf533 a regeneration is done (sensorcommon calcination at corresponding 600 °C with air), 4 and response values and 682 cm−step Another peak to C Fe2H be detected 3O 4 canCO recover theafter original behavior. ThisThe suggests that coke may betested formed on thepresent surfacemore of the electrodes. before and at 1341 cm−1 [36]. Raman spectra of the sample peaks in the − 1 − The iron spinel oxide provides Lewis acidic sites that have a catalytic activity toward aromatics region from 1200 to 1600 cm than the untested sample. Amorphous carbon (1340–1400 cm 1 and 1 [37]) and carbon 1 [38]) are oxidation and [32]. Iron oxide is more by expected aging andtocoke formation 1540–1600 cm−oligomerization nanotubes (1562 cm−affected appear in this leading region. to addition, a reduction of isitsmuch catalytic [33]. Thus, effect sample should isbemeasured. removed This and probably this In there moreactivity fluorescence whencoke the tested may indicate could be dueoftocomplex a surface modification when aromatics are electrode partly oxidized sensor operation. the presence aromatic molecules adsorbed on the surfaceduring and this is in agreement The sensor was exposed to working conditions at 550 °C for 24 days and negligible ageing effect was noticed on the sensor response. Thus, as coke formation is suspected to happen after exposure to aromatics, Raman spectroscopy analyses were performed over the Fe 0.7Cr 1.3O3/8YSZ electrode layer with a gold collector before and after the test in order to check the coke formation. Raman spectroscopy is a technique that
Sensors 2018, 18, x FOR PEER REVIEW
11 of 14
Sensors 2018, 18, 2992
11 of 14
appear in this region. In addition, there is much more fluorescence when the tested sample is measured. This may indicate the presence of complex aromatic molecules adsorbed on the electrode surface is in agreement with the to lack sensitivity afteraromatics exposure(and to the with theand lackthis of sensitivity after exposure theof aforementioned its aforementioned recovery after a aromatics (and itsDespite recovery after a regeneration step). Despite thesenotconsiderations, carbon regeneration step). these considerations, carbon fingerprint is here detected to the limit of fingerprint is here not detected to the limit of this technique. this technique.
Figure 9. Raman spectra of the Fe0.7 Cr1.3 O4 /8YSZ electrode layer with a gold collector after and Figure 9. Raman spectra of the Fe0.7Cr1.3O4/8YSZ electrode layer with a gold collector after and before test. before test.
Regarding NO2 cross cross sensitivity sensitivityon onthe theRE, RE,PtPtisiswell wellknown knownfor foritsits catalytic activity toward Regarding NO catalytic activity toward O2Oat2 2 at mid-high temperatures. Although NO 2 is a common compound in diesel exhaust gas, according to mid-high temperatures. Although NO2 is a common compound in diesel exhaust gas, according to other other potentiometric potentiometric sensors sensors for for selective selective NO NO22 detection, detection, the the catalytic catalytic activity activity of of Pt Pt to to oxygen oxygen redox redox reaction is expected to be much higher than the NO 2 reduction reaction [7,39–41]. Thus, there should reaction is expected to be much higher than the NO2 reduction reaction [7,39–41]. Thus, there should not effect on on the the sensor sensor response, response, although although experimental experimental assays assays are are required. required. not be be aa significant significant effect 4. 4. Conclusions Conclusions Fe0.7 OO3 3was Cr1.3 wasidentified identifiedasasaasuitable suitableworking workingelectrode electrode material material for C22H H44 detection detection in in exhaust exhaust 0.7Cr 1.3 gas streams that contain oxidizable gases together with molecular molecular oxygen. oxygen. The The tested sensor cell comprised comprised a disk-shaped solid oxide-ion conducting conducting electrolyte electrolyte (8YSZ) (8YSZ) coated with the working electrode side. This cellcell assembly Fe0.7Fe Cr 1.3Cr O31.3 /8YSZ || 8YSZ electrode and and aareference referenceelectrode electrode(Pt) (Pt)on onthe theother other side. This assembly O3 /8YSZ || 0.7 ◦ || Pt provides an adequate response to C 2 H 4 , with low cross-sensitivity to CO at 550 °C. This makes 8YSZ || Pt provides an adequate response to C2 H4 , with low cross-sensitivity to CO at 550 C. This the sensor for sensing in conditions like exhaust from diesel cars. The study using wet makes the suitable sensor suitable for sensing in conditions like gases exhaust gases from diesel cars. The study gas streams orstreams adding significant concentration of polycyclic aromatic hydrocarbons (PAHs) revealed using wet gas or adding significant concentration of polycyclic aromatic hydrocarbons (PAHs) the need of further adjusting of surface properties of electrodes to keep high response towards C2H4 revealed the need of further adjusting of surface properties of electrodes to keep high response towards and stability, regardless of the of presence of variable amounts of water or PAHs in ppm C2 Hsignal stability, regardless the presence of variable amounts of water or PAHs inlevel. ppm 4 and signal In theIncase water, the activity toward water gases-shift reaction level. the of case of water, the activity toward water gases-shift reactionshould shouldbebeminimized minimized in in both both working and reference electrodes. working and reference electrodes. following are available online atonline http://www.mdpi.com/1424-8220/18/9/2992/ Supplementary Materials: Materials:The The following are available at www.mdpi.com/xxx/s1, Figure S1: s1, Figure S1:set-up Experimental set-up used measureFigure sensorS2: response, Figure S2: SEMcross-section image of the device Experimental used to measure sensortoresponse, SEM image of the device showing cross-section showing the interface WE-electrolyte corresponding to (a) FeNiO4 , (b) ZnCr O4 , (c) (d)hand, LSC. 4, (b) ZnCr2O4, (c) LSM and (d) 2LSC. OnLSM the and other the interface WE-electrolyte corresponding to (a) FeNiO On the other hand, (e) shows the NiO+5%wt Au surface, Figure S3: Response curves for changes in concentration 2O4 with (e) showsOthewith NiO+5%wt Auand surface, Figure changes concentration ZnCr of ZnCr fresh ink 1 month ink:S3: (a)Response responsecurves to 400 for ppm C2 H4 in with fresh paste,of(b) response to 2 4 fresh ink in and 1 month ink:from (a) response toppm 400 ppm 2H.4 with fresh paste, (b) response to4changes in concentration changes concentration 50 to 200 of C2CH consists of ZnCr2 O as WE and Pt as RE and 4 The sensor at 55050◦ Ctowith a 6%of ofCO2H and with Ar. lower sensor response be observed of2 Thebalanced sensor consists of A ZnCr 2O4 as WE and Pt as can RE and at 550 °Cafter with 1a month 6% of O from 200 ppm 2 4. paste storage.with Thus, the reproducibility does fit the after requirements about reproducibility and balanced Ar. Amaterial lower sensor response can be not observed 1 month of pastestability storage.and Thus, the material with time, Figure S4: not Response to C2about H4 of stability the sensor employing as WE:with (a) LSC and (b) LSC/8YSZ (1:1 reproducibility does fit the transient requirements and reproducibility time, Figure S4: Response vol.), Figure S5: Sensor baseline when exposed to 6% of O2 balanced with argon at 550 ◦ C. The sensor consists of 2H4 of the sensor employing as WE: (a) LSC and (b) LSC/8YSZ (1:1 vol.), Figure S5: Sensor baseline transient Fe Cr to OC /8YSZ as WE and Pt as RE. 0.7
1.3
3
Sensors 2018, 18, 2992
12 of 14
Author Contributions: Conceptualization, J.M.S. and F.T.-R.; Methodology, F.T.-R.; Investigation, J.M.S. and F.T.-R.; Resources, J.M.S.; Data Curation, F.T.-R.; Writing-Original Draft Preparation, J.M.S. and F.T.-R.; Writing-Review & Editing J.M.S. and F.T.-R.; Supervision, J.M.S.; Funding Acquisition, J.M.S. Funding: This research was funded by Spanish government (AP-2003-03478, SEV-2016-0683 and ENE2014-57651). Acknowledgments: Sensata Technologies Holland B.V. is kindly acknowledged. The authors want also acknowledge the Electron Microscopy Service from the Universitat Politècnica de València for their support in the SEM analysis performed in this work. Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.
References 1.
2. 3. 4. 5. 6.
7. 8.
9. 10.
11.
12. 13.
14.
15. 16.
World Health Organization. Health Effects of Transport-Related Air Pollution; Krzyzanowski, M., Kuna-Dibbert, B., Schneider, J., Eds.; World Health Organization, Regional Office for Europe: Denmark, Germany, 2005; ISBN 92 890 1373 7. Emission Standards: Europe: Cars and Light Trucks. Available online: https://dieselnet.com/standards/ eu/ld.php#stds (accessed on 28 June 2018). Sekhar, P.K.; Subramaniyam, K. Electrical characterization of a mixed potential propylene sensor. Sens. Actuators B Chem. 2013, 188, 367–371. [CrossRef] ˙ Re¸sitoglu, ˘ I.A.; Altini¸sik, K.; Keskin, A. The pollutant emissions from diesel-engine vehicles and exhaust aftertreatment systems. Clean Technol. Environ. Policy 2015, 17, 15–27. [CrossRef] Guth, U.; Zosel, J. Electrochemical solid electrolyte gas sensors—Hydrocarbon and NOx analysis in exhaust gases. Ionics 2004, 10, 366–377. [CrossRef] Fadeyev, G.; Kalakin, A.; Demin, A.; Volkov, A.; Brouzgou, A.; Tsiakaras, P. Electrodes for solid electrolyte sensors for the measurement of CO and H2 content in air. Int. J. Hydrogen Energy 2013, 38, 13484–13490. [CrossRef] Yoo, J.; Chatterjee, S.; Wachsman, E.D. Sensing properties and selectivities of a WO3 /YSZ/Pt potentiometric NOx sensor. Sens. Actuators B Chem. 2007, 122, 644–652. [CrossRef] Romanytsia, I.; Viricelle, J.P.; Vernoux, P.; Pijolat, C. Application of advanced morphology Au–X (X = YSZ, ZrO2 ) composites as sensing electrode for solid state mixed-potential exhaust NOx sensor. Sens. Actuators B Chem. 2015, 207, 391–397. [CrossRef] Iio, A.; Ikeda, H.; Anggraini, S.A.; Miura, N. Potentiometric YSZ-based oxygen sensor using BaFeO3 sensing-electrode. Electrochem. Commun. 2014, 48, 134–137. [CrossRef] Plashnitsa, V.V.; Elumalai, P.; Fujio, Y.; Miura, N. Zirconia-based electrochemical gas sensors using nano-structured sensing materials aiming at detection of automotive exhausts. Electrochim. Acta 2009, 54, 6099–6106. [CrossRef] Brosha, E.L.; Mukundan, R.; Brown, D.R.; Garzon, F.H.; Visser, J.H.; Zanini, M.; Zhou, Z.; Logothetis, E.M. CO/HC sensors based on thin films of LaCoO3 and La0.8 Sr0.2 CoO3−δ metal oxides. Sens. Actuators B Chem. 2000, 69, 171–182. [CrossRef] Chevallier, L.; Dibartolomeo, E.; Grilli, M.; Traversa, E. High temperature detection of CO/HCs gases by non-Nernstian planar sensors using Nb2 O5 electrode. Sens. Actuators B Chem. 2008, 130, 514–519. [CrossRef] Yamaguchi, M.; Anggraini, S.A.; Fujio, Y.; Sato, T.; Breedon, M.; Miura, N. Stabilized zirconia-based sensor utilizing SnO2 -based sensing electrode with an integrated Cr2 O3 catalyst layer for sensitive and selective detection of hydrogen. Int. J. Hydrogen Energy 2013, 38, 305–312. [CrossRef] MacAm, E.R.; White, B.M.; Blackburn, B.M.; Di Bartolomeo, E.; Traversa, E.; Wachsman, E.D. La2 CuO4 sensing electrode configuration influence on sensitivity and selectivity for a multifunctional potentiometric gas sensor. Sens. Actuators B Chem. 2011, 160, 957–963. [CrossRef] Fujio, Y.; Sato, T.; Miura, N. Sensing performance of zirconia-based gas sensor using titania sensing-electrode added with palladium. Solid State Ionics 2014, 262, 266–269. [CrossRef] Zosel, J.; Müller, R.; Vashook, V.; Guth, U. Response behavior of perovskites and Au/oxide composites as HC-electrodes in different combustibles. Solid State Ionics 2004, 175, 531–533. [CrossRef]
Sensors 2018, 18, 2992
17.
18.
19. 20. 21. 22.
23.
24.
25. 26.
27. 28. 29.
30. 31. 32. 33. 34. 35.
36.
37.
13 of 14
Mori, M.; Sadaoka, Y.; Nakagawa, S.; Kida, M.; Kojima, T. Development of ethanol and toluene sensing devices with a planar-type structure based on YSZ and modified Pt electrodes. Sens. Actuators B Chem. 2013, 187, 509–513. [CrossRef] Elumalai, P.; Plashnitsa, V.V.; Fujio, Y.; Miura, N. Highly sensitive and selective stabilized zirconia-based mixed-potential-type propene sensor using NiO/Au composite sensing-electrode. Sens. Actuators B Chem. 2010, 144, 215–219. [CrossRef] Fergus, J.W. Sensing mechanism of non-equilibrium solid-electrolyte-based chemical sensors. J. Solid State Electrochem. 2011, 15, 971–984. [CrossRef] Haruta, M. When Gold Is Not Noble: Catalysis by Nanoparticles. Chem. Rec. 2003, 3, 75–87. [CrossRef] [PubMed] Shaikhutdinov, S.K.; Meyer, R.; Naschitzki, M.; Baümer, M.; Freund, H.-J. Size and support effects for CO adsorption on gold model catalysts. Catal. Lett. 2003, 86, 211–219. [CrossRef] De Souza, C.V.; Corr, S.M. Polycyclic aromatic hydrocarbon emissions in diesel exhaust using gas chromatographye-mass spectrometry with programmed temperature vaporization and large volume injection. Atmos. Environ. 2015, 103, 222–230. [CrossRef] Miguel, A.H.; Kirchstetter, T.W.; Harley, R.A.; Hering, S.V. On-Road Emissions of Particulate Polycyclic Aromatic Hydrocarbons and Black Carbon from Gasoline and Diesel Vehicles. Environ. Sci. Technol. 1998, 32, 450–455. [CrossRef] Storey, J.M.; Lewis, S.A.; West, B.H.; Huff, S.P.; Sluder, C.S.; Wagner, R.M.; Domingo, N.; Thomas, J.; Kass, M. Hydrocarbon Species in The Exhaust of Diesel Engines Equipped with Advanced Emissions Control Devices; Final Report CRC Project No. AVFL-10b-2. Available online: www.crcao.org/reports/recentstudies2005/ AVFL-10b-2%20Final%20Report%20January%2031%202005.pdf (accessed on 20 October 2017). Miura, N.; Sato, T.; Anggraini, S.A.; Ikeda, H.; Zhuiykov, S. A review of mixed-potential type zirconia-based gas sensors. Ionics 2014, 20, 901–925. [CrossRef] Striker, T.; Ramaswamy, V.; Armstrong, E.N.; Willson, P.D.; Wachsman, E.D.; Ruud, J.A. Effect of nanocomposite Au–YSZ electrodes on potentiometric sensor response to NOx and CO. Sens. Actuators B. Chem. 2013, 181, 312–318. [CrossRef] Stranzenbach, M.; Saruhan, B. Equivalent circuit analysis on NOx impedance-metric gas sensors. Sens. Actuators B Chem. 2009, 137, 154–163. [CrossRef] Hagen, G.; Dubbe, A.; Fischerauer, G.; Moos, R. Thick-film impedance based hydrocarbon detection based on chromium(III) oxide/zeolite interfaces. Sens. Actuators B Chem. 2006, 118, 73–77. [CrossRef] Miura, N.; Nakatou, M.; Zhuiykov, S. Impedancemetric gas sensor based on zirconia solid electrolyte and oxide sensing electrode for detecting total NOx at high temperature. Sens. Actuators B Chem. 2003, 93, 221–228. [CrossRef] Nakatou, M.; Miura, N. Detection of propene by using new-type impedancemetric zirconia-based sensor attached with oxide sensing-electrode. Sens. Actuators B Chem. 2006, 120, 57–62. [CrossRef] Ikeda, H.; Iio, A.; Anggraini, S.A.; Miura, N. Impedancemetric YSZ-based oxygen sensor using BaFeO3 sensing-electrode. Sens. Actuators B Chem. 2017, 243, 279–282. [CrossRef] Jäger, B.; Wermann, A.; Scholz, P.; Müller, M.; Reislöhner, U.; Stolle, A.; Ondruschka, B. Iron-containing defect-rich mixed metal oxides for Friedel–Crafts alkylation. Appl. Catal. A 2012, 443, 87–95. [CrossRef] Ghorpade, S.P.; Darshane, V.S.; Dixit, S.G. Liquid-phase Friedel-Crafts alkylation using CuCr2−x FexO4 spinel catalysts. Appl. Catal. A 1998, 166, 135–142. [CrossRef] Jadhav, H.S.; Kalubarme, R.S.; Jadhav, A.H.; Seo, J.G. Iron-nickel spinel oxide as an electrocatalyst for non-aqueous rechargeable lithium-oxygen batteries. J. Alloys Compd. 2016, 666, 476–481. [CrossRef] Samarasingha, P.B.; Andersen, N.H.; Sørby, M.H.; Kumar, S.; Nilsen, O.; Fjellvåg, H. Neutron diffraction and Raman analysis of LiMn1.5 Ni0.5 O4 spinel type oxides for use as lithium ion battery cathode and their capacity enhancements. Solid State Ionics 2016, 284, 28–36. [CrossRef] ´ Tian, B.; Swiatowska, J.; Maurice, V.; Zanna, S.; Seyeux, A.; Marcus, P. Binary iron-chromium oxide as negative electrode for lithium-ion micro-batteries—Spectroscopic and microscopic characterization. Appl. Surf. Sci. 2015, 353, 1170–1178. [CrossRef] Veres, M.; Füle, M.; Tóth, S.; Koós, M.; Pócsik, I. Surface enhanced Raman scattering (SERS) investigation of amorphous carbon. Diamond Relat. Mater. 2004, 13, 1412–1415. [CrossRef]
Sensors 2018, 18, 2992
38. 39.
40.
41.
14 of 14
De la Chapelle, M.L.; Lefrant, S.; Journet, C.; Maser, W.; Bernier, P.; Loiseau, A. Raman studies on single walled carbon nanotubes produced by the electric arc technique. Carbon 1998, 36, 705–708. [CrossRef] Dutta, A.; Kaabbuathong, N.; Grilli, M.L.; Di Bartolomeo, E.; Traversa, E. Study of YSZ-Based Electrochemical Sensors with WO3 Electrodes in NO2 and CO Environments. J. Electrochem. Soc. 2003, 150, H33–H37. [CrossRef] Yin, C.; Guan, Y.; Zhu, Z.; Liang, X.; Wang, B.; Diao, Q.; Zhang, H.; Ma, J.; Liu, F.; Sun, Y.; et al. Highly sensitive mixed-potential-type NO2 sensor using porous double-layer YSZ substrate. Sens. Actuators B Chem. 2013, 183, 474–477. [CrossRef] Cai, H.; Sun, R.; Yang, X.; Liang, X.; Wang, C.; Sun, P.; Liu, F.; Zhao, C.; Sun, Y.; Lu, G. Mixed-potential type NOx sensor using stabilized zirconia and MoO3 –In2 O3 nanocomposites. Ceram. Int. 2016, 42, 12503–12507. [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/).
