Article
CO Sensing Performance of a Micro Thermoelectric Gas Sensor with AuPtPd/SnO2 Catalyst and Effects of a Double Catalyst Structure with Pt/α-Al2O3 Tomoyo Goto *,† , Toshio Itoh, Takafumi Akamatsu and Woosuck Shin * Received: 12 November 2015; Accepted: 9 December 2015; Published: 15 December 2015 Academic Editor: W. Rudolf Seitz National Institute of Advanced Industrial Science and Technology (AIST), 2266-98 Shimo-Shidami, Moriyama-ku, Nagoya 463-8560, Japan; [email protected] (T.I.); [email protected] (T.A.) * Correspondence: [email protected] (T.G.); [email protected] (W.S.); Tel.: +81-6-6879-8436 (T.G.); +81-52-736-7107 (W.S.); Fax: +81-6-6879-8439 (T.G.); +81-52-736-7244 (W.S.) † Present address: The Institute of Scientific and Industrial Research (ISIR), Osaka University, 8-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan
Abstract: The CO sensing properties of a micro thermoelectric gas sensor (micro-TGS) with a double AuPtPd/SnO2 and Pt/α-Al2 O3 catalyst were investigated. While several nanometer sized Pt and Pd particles were uniformly dispersed on SnO2 , the Au particles were aggregated as particles measuring >10 nm in diameter. In situ diffuse reflectance Fourier transform Infrared spectroscopy (DRIFT) analysis of the catalyst showed a CO adsorption peak on Pt and Pd, but no clear peak corresponding to the interaction between CO and Au was detected. Up to 200 ˝ C, CO combustion was more temperature dependent than that of H2 , while H2 combustion was activated by repeated exposure to H2 gas during the periodic gas test. Selective CO sensing of the micro-TGS against H2 was attempted using a double catalyst structure with 0.3–30 wt% Pt/α-Al2 O3 as a counterpart combustion catalyst. The sensor output of the micro-TGS decreased with increasing Pt content in the Pt/α-Al2 O3 catalyst, by cancelling out the combustion heat from the AuPtPd/SnO2 catalyst. In addition, the AuPtPd/SnO2 and 0.3 wt% Pt/α-Al2 O3 double catalyst sensor showed good and selective CO detection. We therefore demonstrated that our micro-TGS with double catalyst structure is useful for controlling the gas selectivity of CO against H2 . Keywords: thermoelectric gas sensor; CO oxidation; noble metals; combustion catalyst; gas selectivity
1. Introduction Breath gas contains a range of marker gases that can be associated with disease and metabolism [1,2]. These gases, including CO, H2 , and CH4 , are present in concentrations of several ppm, and hence, accurate analysis of low concentrations of the gases is important for medical applications. In addition, for regular hospital health checks, the gas detector must be easy to adjust and give a fast response. We propose a micro thermoelectric gas sensor (micro-TGS), which uses the thermoelectric detection of a combustion catalyst, for monitoring H2 [3,4], CO [5–7], and CH4 [8,9] in human breath gas. The micro-TGS is an inflammable gas sensor with promising gas responsivity, and shows a clear linear relationship to allow the detection of gases at the ppm level [3]. Au-loaded catalysts have been reported for use in micro-TGSs for CO detection, with Au/Co3 O4 catalysts exhibiting high selectivity to CO but low detection sensitivity [5,6]. Loading of Au, Pt, and Pd on cobalt oxide (Co3 O4 and CoO) catalysts has been reported to improve the detection sensitivity to CO [7], as Pt and Pd particles are also widely used as catalysts for CO oxidation [10,11]. The micro-TGS containing the AuPtPd/CoO catalyst exhibited improved performance at 1 ppm CO in dry air, and
Sensors 2015, 15, 31687–31698; doi:10.3390/s151229873
www.mdpi.com/journal/sensors
Sensors 2015, 15, 31687–31698
good selectivity towards CO over H2 gas [7]. However, a reduction in CO sensitivity of the cobalt oxide catalysts was observed upon continuous use of the micro-TGS [6]. It is assumed that the deactivation of Au particles [12] or cobalt oxide [13] may cause this reduction in CO sensitivity. We therefore chose to focus on SnO2 as a catalyst support for the micro-TGS. SnO2 is a well-known n-type semiconducting oxide, and is used in semiconductor-type gas sensors. In addition, studies into the CO oxidation activity of SnO2 catalysts either with or without noble metals have been previously reported [14–16]. Sakai and Itoh reported the sensing properties of Pt, Pd, and Au loaded on SnO2 (AuPtPd/SnO2 ) thick films as semiconductor-type gas sensors for volatile organic compounds (VOCs). As the addition of three noble metals to SnO2 appears to improve their sensitivity to VOCs [17,18], we attempted the incorporation of an AuPtPd/SnO2 combustion catalyst in our micro-TGS for CO detection. In the micro-TGS systems, the selectivity of the AuPtPd/SnO2 catalyst must be easily controllable. The H2 selectivity for the micro-TGS containing a Pt/α-Al2 O3 catalyst can be controlled by the operating temperature, as the catalyst operating temperature is approximately 100 ˝ C [3,4]. However, for CO and CH4 , selectivity is lower, as a higher catalyst temperature is required for the oxidation of CO and CH4 . In this context, we previously attempted to control CO and CH4 selectivity using a “double catalyst structure” [8,19]. This structure adjusts the balance of the combustion heats of catalysts deposited on the thermoelectric film in the micro-TGS device. Thus, changes in the sensing properties of the micro-TGS are expected when employing the double catalyst structure. We herein report our investigations into the sensing properties of an AuPtPd/SnO2 catalyst on a micro-TGS using a double catalyst structure design. The CO oxidation properties and sensing properties were compared using diffuse reflectance infrared Fourier transform spectroscopy and morphological observations. Furthermore, the effects of the double catalyst structure containing a Pt/α-Al2 O3 catalyst on the selective CO sensing properties of the micro-TGS are also discussed. 2. Experimental Section 2.1. Preparation of the Double Catalyst Micro-TGS The micro-TGS device measuring 4 ˆ 4 mm2 was composed of a p-type B-doped SiGe pattern, a Pt heater, and an electrode line pattern on a double-sided polished Si substrate. Processing details for the B-doped SiGe thin film and its pattern were reported previously [4]. The Pt micro-heater and electrode lines were prepared using a lift-off technique, and the reverse side of the Si substrate was etched out using an aqueous KOH solution to prepare the membrane structure. The AuPtPd/SnO2 catalyst powder was prepared by reference to Itoh’s method [18] and is represented schematically in Figure 1. Au, Pt, and Pd colloids were purchased from Tanaka Kikinzoku Kogyo K.K., Tokyo, Japan, respectively. A catalyst containing 1 wt% Au (2 wt% Au colloidal suspension), 1 wt% Pt (4 wt% Pt colloidal suspension), and 1 wt% Pd (4 wt% Pd colloidal suspension) was added to SnO2 powder (particle size 100 nm; Sigma-Aldrich, St. Louis, MO, USA) in water. Mixing was carried out at 70 ˝ C, followed by drying at 90 ˝ C to give the catalyst powder. This powder was calcined at 350 ˝ C in an electric furnace for 2 h under air. The calcined powder was then dispersed in ethanol, and the floating particles retrieved and mixed with a vehicle containing terpineol, ethyl cellulose, and distilled water (9:1:5 wt. ratio) to give a paste. The catalyst pastes were then applied to the micro-TGS device using an air dispenser (MUSASHI Engineering, Inc., Tokyo, Japan), and the devices were calcined at 300 ˝ C for 2 h. Finally, the devices were mounted on stems, and connected using gold-bonding wire to prepare the micro-TGS.
31688
Sensors 2015, 15, 15, 31687–31698 Sensors 2015, Sensors 2015, page–page 15, page–page
**To Topreparation preparationofof double-catalyst double-catalyststructure structure
SnO SnO 2 powder 2 powder
Pt/a-Al Pt/a-Al 2O 3 3 2O catalyst catalystpaste paste
Pt, and Au,Au, Pt, and Pd Pd colloidal suspension colloidal suspension
PtPt==0.3, 0.3,3,3,3030wt% wt%
Dispensing on Dispensing onSi Sidevice device
AuPtPd/SnO2: Hot side AuPtPd/SnO2: Hot side Pt/a-Al2O3: Cold side
Mixing 70 ºC Mixing 70 ºC
Calcination
Pt/a-Al2O3: Cold side
Calcination
300 ºC, 2h
Calcination
300 ºC, 2h
Calcination350 ºC, 2h
Assembling
350 ºC, 2h
Assembling
Catalyst powder
Catalyst powder
Micro-TGS
Micro-TGS
Characterization
Characterization TEM, STEM-HAADF,
Vehicle
Vehicle
and STEM-EDS observation TEM, STEM-HAADF, DRIFT spectroscopy and STEM-EDS observation DRIFT spectroscopy
Characterization Selectively and stability of Characterization CO, H2 and CH gases sensing Selectively and4 stability of properties CO, H and CH gases sensing
AuPtPd/SnO2 catalyst paste
2
4
properties
AuPtPd/SnO2 catalyst paste
FigureFigure 1. Schematic representation of the powder 2 catalyst 1. Schematic representation of AuPtPd/SnO the AuPtPd/SnO 2 catalyst powderand andmicro microthermoelectric thermoelectric gas sensor (micro-TGS) preparation. gas sensor (micro-TGS) preparation. Figure 1. Schematic representation of the AuPtPd/SnO2 catalyst powder and micro thermoelectric gas sensor (micro-TGS) preparation.
To enhance the micro-TGS selectivity of CO over H2, a double catalyst structure was employed,
To enhance the micro-TGS selectivity of CO over H2 , a double catalyst structure was employed, asTo described in reports [8]. TheofPt/α-Al 2O3 catalyst powder (0.3, 3, and 30 wt%) for H2 enhance theprevious micro-TGS selectivity CO over HO 2, a double catalyst structure was employed, as described in previous reports [8]. The Pt/α-Al powder (0.3, 3, and 30 wt%) 2 3 catalyst combustion was prepared by mixing and subsequent calcination at 300 °C, ˝as wt%) described as described in previous reports [8]. The Pt/α-Al 2O3 catalyst powder (0.3, 3, and 30 for H2 for H2previously combustion was prepared by mixing and subsequent calcination at 300 C, as described [3,4]. combustion was prepared by mixing and subsequent calcination at 300 °C, as described previouslyThe [3,4]. structure and working principle of the double catalyst micro-TGS are based on our previously [3,4]. The structure and working principle of the double catalyst micro-TGS are0.3–30 basedwt% on our previously previously reported study [8,19]. As shown in Figure 2, AuPtPd/SnO 2 and Pt/α-Al 2O3 The structure and working principle of the double catalyst micro-TGS are based on our reported study were [8,19].deposited As shown Figure 2, AuPtPd/SnO 0.3–30 Pt/α-Al catalysts oninthe hot (point A) and cold (point B) wt% side of a micro-TGS device,were 2 and 2 O3 catalysts previously reported study [8,19]. As shown in Figure 2, AuPtPd/SnO2 and 0.3–30 wt% Pt/α-Al2O3 respectively. Hot(point side indicates a position put a catalyst, and cold side showHot theside deposited on the hot A) and cold (pointtoB)commonly side of a micro-TGS device, respectively. catalysts were deposited on the hot (point A) and cold (point B) side of a micro-TGS device, position of without catalyst on micro-TGS. It is expected that the AuPtPd/SnO 2 catalyst will oxidize indicates a position to commonly put a catalyst, and cold side show the position of without catalyst respectively. Hot side indicates a position to commonly put a catalyst, and cold side show the both H2 and However, Pt/α-Al 2O3 is a combustion catalyst for H2, [3,4] hence, the combustion on position micro-TGS. It isCO. expected that the AuPtPd/SnO both2H However, 2 catalyst 2 and CO. of without catalyst on micro-TGS. It is expected thatwill the oxidize AuPtPd/SnO catalyst oxidize heat of the Pt/α-Al2O3 catalyst (at point B) reduces the temperature difference between thewill points A Pt/α-Al O is a combustion catalyst for H , [3,4] hence, the combustion heat of the Pt/α-Al O 2 2 3and CO. However, Pt/α-Al2O3 is 2 a combustion catalyst for H2, [3,4] hence, the combustion 2 3 catalyst both H and B. As a result, when a mixture of H2 and CO is introduced into the calorimetric-TGS device, the (at heat pointofB)the reduces the difference between the points A and B. As a result, when a mixture Pt/α-Al 2of Otemperature 3the catalyst (at point B) reduces the points sensor response AuPtPd/SnO 2 catalyst to H2the willtemperature be inhibited.difference The designbetween and composition of A of H and CO is introduced into the calorimetric-TGS device, the sensor response of the AuPtPd/SnO 2 single and B. Asand a result, when a mixture of H2 andofCO introduced2 into the calorimetric-TGS the 2 double catalyst type structures the is AuPtPd/SnO and 0.3–30 wt% Pt/α-Al2O3device, catalysts catalyst toresponse H inhibited. The design and composition of single and double catalyst type sensor the AuPtPd/SnO 2 catalyst to H2 1. will be inhibited. The design and composition of 2 willofbe of the micro-TGSs are given in Figure 3 and Table structures of the AuPtPd/SnO 0.3–30 wt% of the micro-TGSs given in single and double catalyst type structures of thePt/α-Al AuPtPd/SnO 2 and 0.3–30 wt% Pt/α-Al2Oare 3 catalysts 2 and 2 O3 catalysts Figure 3 and Table 1.are given in Figure 3 and Table 1. of the micro-TGSs
(a)
(a)
(b)
Pt line Heater
Pt line
A: Hot side (AuPtPd/SnO2)
CO or H2 + O2
(b)
B: Cold side Thermoelectric (Pt/a-Al2O3)
SiGe line
2H2+O2
+ O2
2H2O
Oxidation
Oxidation
CO or H2 A: Hot side
B: Cold side Heater (Pt/a-Al2O3)
A: Hot side A: Hot side Si membrane (AuPtPd/SnO2)
CO2 or H2O
CO2 or H2O
2H2+O2
Oxidation
Oxidation Thermoelectric
1 wt%Au, 1 wt%Pt, 1 wt%Pd /SnO2 catalyst
B: Cold side
2H2O
B: Cold side
30-0.3 wt% Pt /a-Al2O3 catalysts
Thermoelectric
Si membrane
Thermoelectric 1 wt%Au, 1 wt%Pt, 1 wt%Pd 30-0.3 wt% Pt SiGe line /a-Al2O3 catalysts Figure 2. Optical images of the AuPtPd/SnO2 and/SnO Pt/Al 2O3 catalysts on the Si membrane of the 2 catalyst micro-TGS (a), and a side view of the double catalyst structure (b).
Figure Optical imagesofof the AuPtPd/SnO2 and and Pt/Al2OO 3 catalysts on the Si membrane of the 3 Pt/Al Figure 2. 2. Optical images the AuPtPd/SnO 2 2 3 catalysts on the Si membrane of the micro-TGS (a), and a side view of the double catalyst structure (b). micro-TGS (a), and a side view of the double catalyst structure (b).
3
31689
Sensors 2015, 15, 31687–31698
Sensors 2015, 15, page–page
0.3 wt%
(b)
(a)
A
A
B 3 wt%
(c)
A
B 30 wt%
(d)
A
B
B
Figure 3. Double catalyst AuPtPd/SnO and 0.3–30 wt% Pt/α-Al Figure 3. Double catalyststructures structures with with AuPtPd/SnO 2 (A: HotHot side)side) and 0.3–30 wt% Pt/α-Al 2O3 (B: 2 O3 2 (A: (B: Cold themicro-TGS: micro-TGS:(a)(a)No No catalyst; Pt catalyst; (c) 3 Pt wt% Pt catalyst, Coldside) side)catalysts catalysts for the PtPt catalyst; (b)(b) 0.3 0.3 wt%wt% Pt catalyst; (c) 3 wt% catalyst, and (d) 30 wt% Pt catalyst. and (d) 30 wt% Pt catalyst. Table 1. Design of AuPtPd/SnO2 and and Pt/α-Al 2O3 double catalyst structures. Table 1. Design of AuPtPd/SnO Pt/α-Al 2 2 O3 double catalyst structures. Pt content in Pt/α-Al2O3
Notation
Design
0Pt
AuPtPd/SnO2 + no catalyst
Design
Notation 0Pt 0.3Pt 0.3Pt 3Pt 3Pt 30Pt 30Pt
Pt content in Pt/α-Al2 O3 wt% wt%
AuPtPd/SnO2 + no catalyst AuPtPd/SnO 2 + Pt/α-Al2O3 AuPtPd/SnO 2 + Pt/α-Al2 O3 AuPtPd/SnO AuPtPd/SnO2 + Pt/α-Al 2O 3 2 + Pt/α-Al 2 O3 AuPtPd/SnO2 + Pt/α-Al2 O3 AuPtPd/SnO2 + Pt/α-Al2O3
0
0
0.3 0.3 3 3 30
30
2.2. Catalyst Powder Characterization 2.2. Catalyst Powder Characterization morphology thecatalyst catalystpowder powder was electron microscopy The The morphology of of the was observed observedbybytransmission transmission electron microscopy (TEM; Tecnai Osiris, FEI, OR, USA) with high-angle annular dark field scanning transmission (TEM; Tecnai Osiris, FEI, OR, USA) with high-angle annular dark field scanning transmission electron electron microscopy (STEM-HAADF), at an accelerating 200Elemental kV. Elemental mapping microscopy (STEM-HAADF), at an accelerating voltagevoltage of 200ofkV. mapping of of Au, Pt, Au, Pt, Pd, and Sn was carried out using STEM-energy dispersive X-ray spectroscopy (STEM-EDS). Pd, and Sn was carried out using STEM-energy dispersive X-ray spectroscopy (STEM-EDS). CO CO adsorption and oxidation of the catalyst powder were investigated by diffuse reflectance Fourier adsorption and oxidation of the catalyst powder were investigated by diffuse reflectance Fourier transform Infrared spectroscopy (DRIFT) analysis (Nexus 470 FTIR, Nicolet, Waltham, MA, USA), at transform spectroscopy (DRIFT) analysis 470 FTIR, Nicolet, Waltham, MA, at 25, 50, Infrared 100, 200, and 300 °C. As measurement gases,(Nexus 10,000 ppm CO in dry air and 10,000 ppm COUSA), in ˝ 25, 50, and As of measurement gases, 10,000 in dryin airabsorbance and 10,000 ppm CO Ar100, were200, used at 300 a flowC.rate 50 cm3/min. DRIFT spectra ppm were CO measured mode 3 in Arwithout were used at a flow 50orcm switching carrierrate gas of (Air Ar)./min. DRIFT spectra were measured in absorbance mode
without switching carrier gas (Air or Ar). 2.3. Investigation of Gas Sensing Properties
2.3. Investigation of Gas Sensing Properties
Measurement system of sensing properties of micro-TGS is shown in Figure 4. The sensor response of micro-TGS 2, CO, and CH4 wasofinvestigated a gas flow chamber a Measurement systemfor of H sensing properties micro-TGSusing is shown in Figure 4. with The sensor volume of 60 cm3. The to the micro-TGS device wasaadjusted tochamber control the heater response of micro-TGS for voltage H2 , CO,applied and CH was investigated using gas flow with a volume 4 (50–250 °C) and was monitored using an IR camera (LAIRD-270A, Nikon Co., Tokyo, of 60temperatures cm3 . The voltage applied to the micro-TGS device was adjusted to control the heater temperatures Japan). To test the micro-TGS, pure dry air was allowed to flow into the chamber together with the (50–250 ˝ C) and was monitored using an IR camera (LAIRD-270A, Nikon Co., Tokyo, Japan). To test test gases (in dry air) in the following order: Dry air, test gas, and dry air. Each step lasted 90 s with a the micro-TGS, pure dry air was allowed to flow into the chamber together with the test gases (in dry gas flow rate of 200 cm3/min. The voltage across the thermoelectric thin film of the micro-TGS varied air) in the following air, test gas, and dry air. Each stepSeebeck lasted principle. 90 s withThe a gas flow rate by the combustionorder: heat ofDry the catalysts, a phenomenon based on the voltage 3 /min. The voltage across the thermoelectric thin film of the micro-TGS varied by the of 200 cm signal (Vs) was corrected using a digital multimeter (K2700, TFF Co. Keithley Inst., Tokyo, Japan) combustion of the catalysts, a phenomenon based onwith the 10,000 Seebeck principle. voltage followingheat pretreatment of the double catalyst micro-TGS ppm H2 in dryThe air at 200 °C.signal The response of using Vs (ΔV) is defined by Equation (1): TFF Co. Keithley Inst., Tokyo, Japan) following (Vs) was corrected a digital multimeter (K2700,
pretreatment of the double catalyst micro-TGS with 10,000 ppm H2 in dry air at 200 ˝ C. The response 4 of Vs (∆V) is defined by Equation (1): ∆V “ Vsp170´180q ´ Vsp80´90q 31690
(1)
Sensors 2015, 15, page–page
ΔV = Vs(170–180) – Vs(80–90)
Sensors 2015, 15, page–page
(1)
Sensors 15, 31687–31698 where2015, Vs(170–180) and Vs(80–90)
are the averages of Vs over 170–180 s (in test gas) and 80–90 s (in dry air), respectively. The sensor response was investigated the voltage across the micro-TGS ΔV = Vs(170–180) by – Vsmeasuring (80–90) (1) at 1000 ppm CO, H2, and CH4 at 50–250 °C and 10–10,000 ppm CO or H2 at 200 °C. The stability of the where Vs(170–180) andVs Vsp(80–90) thethe averages of Vsofover 170–180 s (in test gas) andgas) 80–90 s (in dry air), where Vsp170–180 averages Vs over 170–180 s (in test and 80–90 s (in dry q and 80–90are q are sensor responses was determined by subjecting the micro-TGS to a gas sensitivity test at 200 °C, for respectively. The was investigated by by measuring the voltage across the micro-TGS air), respectively. Thesensor sensorresponse response was investigated measuring the voltage across the micro-TGS 1, 2, at and 9 days.CO, H2, and CH4 at 50–250˝°C and 10–10,000 ppm CO or H2 at 200 °C. The 1000 of theof the at 1000 ppmppm CO, H2 , and CH4 at 50–250 C and 10–10,000 ppm CO or H2 at 200 ˝ C. stability The stability sensor responses was determined by subjecting the micro-TGS to a gas sensitivity test at 200 °C,˝for sensor responses was determined by subjecting the micro-TGS to a gas sensitivity test at 200 C, for 1, 1, 2, and 9 days. 2, and 9 days. Control of the gas flow rate and switching of gas type Control of the gas flowSignal rate and switching of gas type from multimeter
PC
Signal fromsensor multimeter Image Signal from
Digital multimeter Digital multimeter
Gas flow controller
Gas tank
CCD camera
CCD camera
Gas flow controller
Gas tank
PC
Image
Signal from sensor
Gas flow Micro-TGS
Gas flow
Micro-TGS
Measuring
Waste gas Waste gas
Measuring chamber chamber
Stabilized power
Stabilized supply power supply
Gas supply Gas supply
Figure 4. of gas sensing properties of micro-TGS. micro-TGS. Figure 4. Measurement Measurement system of gas gassensing sensingproperties properties of Figure 4. Measurementsystem system of of micro-TGS.
3. Results and Discussion 3. Results and Discussion 3. Results and Discussion 3.1. Catalyst Microstructure 3.1. Catalyst Microstructure Figure 5 shows the TEM images of the AuPtPd/SnO2 catalyst powder. Both well-dispersed Figure showsthe theTEM TEM images of the AuPtPd/SnO 2 catalyst powder. well-dispersed 55shows images of the AuPtPd/SnO powder. Both Both well-dispersed SnO2 2 catalyst SnO2 particles (20–50 nm) (Figure 5a), and small metal particles (3 nm) supported on SnO2 particles SnO2 particles (20–50 nm) (Figure 5a),small and small metal particles (3 supported nm) supported on 2SnO 2 particles particles (20–50 nm) (Figure 5a), and metal particles (3 nm) on SnO particles can clearly be seen (Figure 5b). It should be noted here that the high combustion performance of the can can clearly be seen (Figure 5b). It should notedhere herethat that The thehigh high combustion performance of the clearly be seen (Figure It should bebe noted performance catalyst is affected by5b). its porous structure and particle size.the graincombustion size of the aggregated metalof the catalyst is affected by its porous structure and particle size. The grain size of the aggregated metal catalyst is affected by itsnanometers, porous structure particle size. The grain sizeresulting of the aggregated particles was several similarand to the starting colloidal particle, in relativelymetal particles was severalnanometers, nanometers, similar the starting colloidal particle, resulting inAu) relatively particles several similar to to the starting colloidal particle, resulting relatively goodwas dispersion. However, to investigate whether each metal element (i.e., Pt, Pd,inand is good good dispersion. However, to investigate whether each metal element (i.e., Pt, Pd, and is alloyed or not, the catalyst was analyzed by STEM and elemental mapping. dispersion. However, to investigate whether each metal element (i.e., Pt, Pd, and Au) is alloyed Au) or not,
alloyed or not, catalystby was analyzed by STEMmapping. and elemental mapping. the catalyst wasthe analyzed STEM and elemental
Figure 5. Transmission electron microscopy (TEM) images of the AuPtPd/SnO2 catalyst powder: (a) low magnification (17,500) and (b) high magnification (88,000).
5 5. Transmission electron microscopy (TEM) (TEM) images images of of the the AuPtPd/SnO AuPtPd/SnO22 catalyst Figure 5. catalyst powder: powder: (17,500) and and (b) (b) high high magnification magnification (88,000). (88,000). (a) low magnification (17,500)
5
31691
Sensors 2015, 15, 31687–31698 Sensors 2015, 15, page–page
Figure 6 shows the High-angle annular dark field scanning transmission electron microscopy Figure 6images shows the annular dark field scanning transmission electron microscopy (STEM-HAADF) andHigh-angle STEM-energy dispersive X-ray spectroscopy (STEM-EDS) mapping of the (STEM-HAADF) images and STEM-energy dispersive X-ray spectroscopy (STEM-EDS) mapping of Au, Pt, Pd, and Sn distributions in the catalyst powder. Low magnification images (Figure 6a) indicate the Au, Pt, Pd, and Sn distributions in the catalyst powder. Low magnification images (Figure 6a) that the noble metals were dispersed over the SnO2 particles. However, in the high magnification indicate that the noble metals were dispersed over the SnO2 particles. However, in the high images (Figure 6b), only the Pt and Pd nanoparticles were deposited and dispersed on SnO2 particles, magnification images (Figure 6b), only the Pt and Pd nanoparticles were deposited and dispersed on while the Au nanoparticles aggregated as 10 nm particles, overlapping with Pt and Pd, as indicated by SnO2 particles, while the Au nanoparticles aggregated as 10 nm particles, overlapping with Pt and arrows overlap 6b. image of Figure 6b. Pd,in asthe indicated byimage arrowsof inFigure the overlap
Figure 6. High-angle annular dark field scanning transmission electron microscopy (STEM-HAADF)
Figure 6. High-angle annular dark field scanning transmission electron microscopy (STEM-HAADF) and STEM-energy dispersive X-ray spectroscopy (STEM-EDS) images of the AuPtPd/SnO2 catalyst and STEM-energy dispersive image X-ray and spectroscopy images of the AuPtPd/SnO catalyst powder. (a) STEM-HAADF EDS maps (STEM-EDS) of Au, Pt, Pd, Sn, and Overlap (Au-Pt-Pd) at2low powder. (a) STEM-HAADF image and EDS maps of Au, Pt, Pd, Sn, and Overlap (Au-Pt-Pd) at low magnification (110,000), and (b) STEM-HAADF image and EDS maps of Au, Pt, Pd, Sn, and Overlap magnification (110,000), and (b) STEM-HAADF image and EDS maps of Au, Pt, Pd, Sn, and Overlap at at high magnification (630,000). high magnification (630,000). 3.2. DRIFT Characterization of the Catalyst
3.2. DRIFT Characterization the Catalyst Figure 7 shows the of 1000–4000 cm−1 region of the diffuse reflectance Fourier transform Infrared spectroscopy (DRIFT) spectra of the AuPtPd/SnO2 catalyst at 25–300 °C in 10,000 ppm CO in both dry
Figure 7 shows the 1000–4000 cm´1 region of the diffuse reflectance Fourier transform Infrared air (Figure 7a) and in Ar (Figure 7b). As can be seen in Figure 7a, at lower temperatures (25–100 °C), spectroscopy (DRIFT) spectra of the AuPtPd/SnO at 25–300 ˝ C in 10,000 ppm CO in both dry 2 catalyst bands at 1082, 1221, 1311, and 1551 cm−1 were observed, corresponding to vibrations of the ˝ air (Figure 7a) species and in Ar (Figure 7b). As can2be seen in Figure lower temperatures (25–100 carbonate on noble metals or SnO [15,20]. The bands7a, at at 1834, 2078, 2116, and 2175 cm−1 C), ´ 1 bandscould at 1082, 1221, 1311, and cm the were observed, to broad vibrations of 1834 the carbonate be assigned to CO. In1551 particular, sharp peak at corresponding 2078 cm−1 and the peak at cm−1 ´ 1 species on nobleto metals or SnO bands at 1834, 2116, and 2175respectively cm could be assigned correspond the linear and bridgingThe bonds of the CO-Pt2078, and CO-Pd bands, [21]. The 2 [15,20]. 1 and the 1 correspond to peak corresponding to COat adsorption decreased with increasing the1834 catalyst to CO. In intensity particular, the sharp peak 2078 cm´ broad peak at cm´temperature. At catalyst temperatures of of 200the andCO-Pt 300 °C, bands at 2314 andrespectively 2359 cm−1, corresponding to CO 2, the linear and bridging bonds and CO-Pd bands, [21]. The peak intensity were observed, confirming the oxidation of CO by the AuPtPd/SnO 2 catalyst. For the experiments corresponding to CO adsorption decreased with increasing the catalyst temperature. At catalyst employing 10,000 ppm CO in Ar (Figure 7b), the peaks corresponding to the carbonate species (at temperatures of 200 and 300 ˝ C, bands at 2314 and 2359 cm´1 , corresponding to CO2 , were observed, 1097, 1225, 1313, 1549, 2079, and 2172 cm−1) were detected at 25–200 °C. However, in contrast to the confirming the oxidation of CO by the AuPtPd/SnO2 catalyst. For the experiments employing experiments employing 10,000 ppm CO in dry air (Figure 7a), peaks corresponding to CO2 were 10,000 ppm CO in Ar (Figure 7b), the peaks corresponding to the carbonate species (at 1097, 1225, not detected. 1313, 1549, 2079, and 2172 cm´1 ) were detected at 25–200 ˝ C. However, in contrast to the experiments employing 10,000 ppm CO in dry air (Figure 7a), peaks corresponding to CO2 were not detected.
6
31692
Sensors 2015, 15, 31687–31698
Sensors 2015, 15, page–page
100 ºC 50 ºC
1549
1313 1225 1097
2079
2172
2357
2550
300 ºC 1834
(a.u.) Absorbance/ a.u. Absorbance
1221 1082
1834
200 ºC
3691
1573
1446 1311 1171 1128
2359 2314 2175 2116 2078
300 ºC 1551
Absorbance / a.u. Absorbance (a.u.)
2553
(b) 3701
(a)
200 ºC 100 ºC 50 ºC 25 ºC
25 ºC
4000
3000 2000 -1) Wavenumber /(cm cm -1 Wavenumber
4000
1000
3000 2000 Wavenumber Wavenumber /(cm cm -1-1)
1000
FigureFigure 7. Diffuse reflectance Infrared spectroscopy spectroscopy (DRIFT) spectra of the 7. Diffuse reflectanceFourier Fourier transform transform Infrared (DRIFT) spectra of the ˝ C, in (a) 10,000 ppm CO in dry air, and (b) 10,000 ppm CO AuPtPd/SnO catalyst powder at 25–300 2 AuPtPd/SnO 2 catalyst powder at 25–300 °C, in (a) 10,000 ppm CO in dry air, and (b) 10,000 ppm CO in Ar.in Ar.
The DRIFT spectra (Figure 7) clearly showed a peak corresponding to CO adsorption on the Pt
The DRIFT spectra (Figure 7) clearly showed a peak corresponding to CO adsorption on the and Pd particles at 2078 cm−1, but no sharp peak corresponding to CO adsorption on the metallic Au Pt and Pd particles at 2078 cm´−11 , but no sharp peak corresponding to CO adsorption on the metallic site was detected at 2112 cm as shown in previous researches [15,22]. This indicated that less CO ´1 as shown in previous researches [15,22]. This indicated that less Au site was detected at 2112 cmAu oxidation took place on the particles compared to the Pt and Pd particles. As shown in the CO oxidation took place on the6),Au to the Ptthus andincreasing Pd particles. shown in the STEM-EDS images (Figure Auparticles particles compared formed aggregates, theirAs particle size STEM-EDS images (Figure 6), Au particles formed aggregates, thus increasing their particle (≥10 nm). This can be explained by previous studies showing that the oxidation reaction of the Au size catalyst slowed with increasingby particle diameter, with a particlethat sizethe of 30Pt. these results, 0Pt > 0.3Pt ~ 3Pt > 30Pt. Considering these results, an optimal Pt content of 0.3Pt2O~3 3Pt > 30Pt. Considering these results, Pt content of
CO Sensing Performance of a Micro Thermoelectric Gas Sensor with AuPtPd/SnO2 Catalyst and Effects of a Double Catalyst Structure with Pt/α-Al2O3 Tomoyo Goto *,† , Toshio Itoh, Takafumi Akamatsu and Woosuck Shin * Received: 12 November 2015; Accepted: 9 December 2015; Published: 15 December 2015 Academic Editor: W. Rudolf Seitz National Institute of Advanced Industrial Science and Technology (AIST), 2266-98 Shimo-Shidami, Moriyama-ku, Nagoya 463-8560, Japan; [email protected] (T.I.); [email protected] (T.A.) * Correspondence: [email protected] (T.G.); [email protected] (W.S.); Tel.: +81-6-6879-8436 (T.G.); +81-52-736-7107 (W.S.); Fax: +81-6-6879-8439 (T.G.); +81-52-736-7244 (W.S.) † Present address: The Institute of Scientific and Industrial Research (ISIR), Osaka University, 8-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan
Abstract: The CO sensing properties of a micro thermoelectric gas sensor (micro-TGS) with a double AuPtPd/SnO2 and Pt/α-Al2 O3 catalyst were investigated. While several nanometer sized Pt and Pd particles were uniformly dispersed on SnO2 , the Au particles were aggregated as particles measuring >10 nm in diameter. In situ diffuse reflectance Fourier transform Infrared spectroscopy (DRIFT) analysis of the catalyst showed a CO adsorption peak on Pt and Pd, but no clear peak corresponding to the interaction between CO and Au was detected. Up to 200 ˝ C, CO combustion was more temperature dependent than that of H2 , while H2 combustion was activated by repeated exposure to H2 gas during the periodic gas test. Selective CO sensing of the micro-TGS against H2 was attempted using a double catalyst structure with 0.3–30 wt% Pt/α-Al2 O3 as a counterpart combustion catalyst. The sensor output of the micro-TGS decreased with increasing Pt content in the Pt/α-Al2 O3 catalyst, by cancelling out the combustion heat from the AuPtPd/SnO2 catalyst. In addition, the AuPtPd/SnO2 and 0.3 wt% Pt/α-Al2 O3 double catalyst sensor showed good and selective CO detection. We therefore demonstrated that our micro-TGS with double catalyst structure is useful for controlling the gas selectivity of CO against H2 . Keywords: thermoelectric gas sensor; CO oxidation; noble metals; combustion catalyst; gas selectivity
1. Introduction Breath gas contains a range of marker gases that can be associated with disease and metabolism [1,2]. These gases, including CO, H2 , and CH4 , are present in concentrations of several ppm, and hence, accurate analysis of low concentrations of the gases is important for medical applications. In addition, for regular hospital health checks, the gas detector must be easy to adjust and give a fast response. We propose a micro thermoelectric gas sensor (micro-TGS), which uses the thermoelectric detection of a combustion catalyst, for monitoring H2 [3,4], CO [5–7], and CH4 [8,9] in human breath gas. The micro-TGS is an inflammable gas sensor with promising gas responsivity, and shows a clear linear relationship to allow the detection of gases at the ppm level [3]. Au-loaded catalysts have been reported for use in micro-TGSs for CO detection, with Au/Co3 O4 catalysts exhibiting high selectivity to CO but low detection sensitivity [5,6]. Loading of Au, Pt, and Pd on cobalt oxide (Co3 O4 and CoO) catalysts has been reported to improve the detection sensitivity to CO [7], as Pt and Pd particles are also widely used as catalysts for CO oxidation [10,11]. The micro-TGS containing the AuPtPd/CoO catalyst exhibited improved performance at 1 ppm CO in dry air, and
Sensors 2015, 15, 31687–31698; doi:10.3390/s151229873
www.mdpi.com/journal/sensors
Sensors 2015, 15, 31687–31698
good selectivity towards CO over H2 gas [7]. However, a reduction in CO sensitivity of the cobalt oxide catalysts was observed upon continuous use of the micro-TGS [6]. It is assumed that the deactivation of Au particles [12] or cobalt oxide [13] may cause this reduction in CO sensitivity. We therefore chose to focus on SnO2 as a catalyst support for the micro-TGS. SnO2 is a well-known n-type semiconducting oxide, and is used in semiconductor-type gas sensors. In addition, studies into the CO oxidation activity of SnO2 catalysts either with or without noble metals have been previously reported [14–16]. Sakai and Itoh reported the sensing properties of Pt, Pd, and Au loaded on SnO2 (AuPtPd/SnO2 ) thick films as semiconductor-type gas sensors for volatile organic compounds (VOCs). As the addition of three noble metals to SnO2 appears to improve their sensitivity to VOCs [17,18], we attempted the incorporation of an AuPtPd/SnO2 combustion catalyst in our micro-TGS for CO detection. In the micro-TGS systems, the selectivity of the AuPtPd/SnO2 catalyst must be easily controllable. The H2 selectivity for the micro-TGS containing a Pt/α-Al2 O3 catalyst can be controlled by the operating temperature, as the catalyst operating temperature is approximately 100 ˝ C [3,4]. However, for CO and CH4 , selectivity is lower, as a higher catalyst temperature is required for the oxidation of CO and CH4 . In this context, we previously attempted to control CO and CH4 selectivity using a “double catalyst structure” [8,19]. This structure adjusts the balance of the combustion heats of catalysts deposited on the thermoelectric film in the micro-TGS device. Thus, changes in the sensing properties of the micro-TGS are expected when employing the double catalyst structure. We herein report our investigations into the sensing properties of an AuPtPd/SnO2 catalyst on a micro-TGS using a double catalyst structure design. The CO oxidation properties and sensing properties were compared using diffuse reflectance infrared Fourier transform spectroscopy and morphological observations. Furthermore, the effects of the double catalyst structure containing a Pt/α-Al2 O3 catalyst on the selective CO sensing properties of the micro-TGS are also discussed. 2. Experimental Section 2.1. Preparation of the Double Catalyst Micro-TGS The micro-TGS device measuring 4 ˆ 4 mm2 was composed of a p-type B-doped SiGe pattern, a Pt heater, and an electrode line pattern on a double-sided polished Si substrate. Processing details for the B-doped SiGe thin film and its pattern were reported previously [4]. The Pt micro-heater and electrode lines were prepared using a lift-off technique, and the reverse side of the Si substrate was etched out using an aqueous KOH solution to prepare the membrane structure. The AuPtPd/SnO2 catalyst powder was prepared by reference to Itoh’s method [18] and is represented schematically in Figure 1. Au, Pt, and Pd colloids were purchased from Tanaka Kikinzoku Kogyo K.K., Tokyo, Japan, respectively. A catalyst containing 1 wt% Au (2 wt% Au colloidal suspension), 1 wt% Pt (4 wt% Pt colloidal suspension), and 1 wt% Pd (4 wt% Pd colloidal suspension) was added to SnO2 powder (particle size 100 nm; Sigma-Aldrich, St. Louis, MO, USA) in water. Mixing was carried out at 70 ˝ C, followed by drying at 90 ˝ C to give the catalyst powder. This powder was calcined at 350 ˝ C in an electric furnace for 2 h under air. The calcined powder was then dispersed in ethanol, and the floating particles retrieved and mixed with a vehicle containing terpineol, ethyl cellulose, and distilled water (9:1:5 wt. ratio) to give a paste. The catalyst pastes were then applied to the micro-TGS device using an air dispenser (MUSASHI Engineering, Inc., Tokyo, Japan), and the devices were calcined at 300 ˝ C for 2 h. Finally, the devices were mounted on stems, and connected using gold-bonding wire to prepare the micro-TGS.
31688
Sensors 2015, 15, 15, 31687–31698 Sensors 2015, Sensors 2015, page–page 15, page–page
**To Topreparation preparationofof double-catalyst double-catalyststructure structure
SnO SnO 2 powder 2 powder
Pt/a-Al Pt/a-Al 2O 3 3 2O catalyst catalystpaste paste
Pt, and Au,Au, Pt, and Pd Pd colloidal suspension colloidal suspension
PtPt==0.3, 0.3,3,3,3030wt% wt%
Dispensing on Dispensing onSi Sidevice device
AuPtPd/SnO2: Hot side AuPtPd/SnO2: Hot side Pt/a-Al2O3: Cold side
Mixing 70 ºC Mixing 70 ºC
Calcination
Pt/a-Al2O3: Cold side
Calcination
300 ºC, 2h
Calcination
300 ºC, 2h
Calcination350 ºC, 2h
Assembling
350 ºC, 2h
Assembling
Catalyst powder
Catalyst powder
Micro-TGS
Micro-TGS
Characterization
Characterization TEM, STEM-HAADF,
Vehicle
Vehicle
and STEM-EDS observation TEM, STEM-HAADF, DRIFT spectroscopy and STEM-EDS observation DRIFT spectroscopy
Characterization Selectively and stability of Characterization CO, H2 and CH gases sensing Selectively and4 stability of properties CO, H and CH gases sensing
AuPtPd/SnO2 catalyst paste
2
4
properties
AuPtPd/SnO2 catalyst paste
FigureFigure 1. Schematic representation of the powder 2 catalyst 1. Schematic representation of AuPtPd/SnO the AuPtPd/SnO 2 catalyst powderand andmicro microthermoelectric thermoelectric gas sensor (micro-TGS) preparation. gas sensor (micro-TGS) preparation. Figure 1. Schematic representation of the AuPtPd/SnO2 catalyst powder and micro thermoelectric gas sensor (micro-TGS) preparation.
To enhance the micro-TGS selectivity of CO over H2, a double catalyst structure was employed,
To enhance the micro-TGS selectivity of CO over H2 , a double catalyst structure was employed, asTo described in reports [8]. TheofPt/α-Al 2O3 catalyst powder (0.3, 3, and 30 wt%) for H2 enhance theprevious micro-TGS selectivity CO over HO 2, a double catalyst structure was employed, as described in previous reports [8]. The Pt/α-Al powder (0.3, 3, and 30 wt%) 2 3 catalyst combustion was prepared by mixing and subsequent calcination at 300 °C, ˝as wt%) described as described in previous reports [8]. The Pt/α-Al 2O3 catalyst powder (0.3, 3, and 30 for H2 for H2previously combustion was prepared by mixing and subsequent calcination at 300 C, as described [3,4]. combustion was prepared by mixing and subsequent calcination at 300 °C, as described previouslyThe [3,4]. structure and working principle of the double catalyst micro-TGS are based on our previously [3,4]. The structure and working principle of the double catalyst micro-TGS are0.3–30 basedwt% on our previously previously reported study [8,19]. As shown in Figure 2, AuPtPd/SnO 2 and Pt/α-Al 2O3 The structure and working principle of the double catalyst micro-TGS are based on our reported study were [8,19].deposited As shown Figure 2, AuPtPd/SnO 0.3–30 Pt/α-Al catalysts oninthe hot (point A) and cold (point B) wt% side of a micro-TGS device,were 2 and 2 O3 catalysts previously reported study [8,19]. As shown in Figure 2, AuPtPd/SnO2 and 0.3–30 wt% Pt/α-Al2O3 respectively. Hot(point side indicates a position put a catalyst, and cold side showHot theside deposited on the hot A) and cold (pointtoB)commonly side of a micro-TGS device, respectively. catalysts were deposited on the hot (point A) and cold (point B) side of a micro-TGS device, position of without catalyst on micro-TGS. It is expected that the AuPtPd/SnO 2 catalyst will oxidize indicates a position to commonly put a catalyst, and cold side show the position of without catalyst respectively. Hot side indicates a position to commonly put a catalyst, and cold side show the both H2 and However, Pt/α-Al 2O3 is a combustion catalyst for H2, [3,4] hence, the combustion on position micro-TGS. It isCO. expected that the AuPtPd/SnO both2H However, 2 catalyst 2 and CO. of without catalyst on micro-TGS. It is expected thatwill the oxidize AuPtPd/SnO catalyst oxidize heat of the Pt/α-Al2O3 catalyst (at point B) reduces the temperature difference between thewill points A Pt/α-Al O is a combustion catalyst for H , [3,4] hence, the combustion heat of the Pt/α-Al O 2 2 3and CO. However, Pt/α-Al2O3 is 2 a combustion catalyst for H2, [3,4] hence, the combustion 2 3 catalyst both H and B. As a result, when a mixture of H2 and CO is introduced into the calorimetric-TGS device, the (at heat pointofB)the reduces the difference between the points A and B. As a result, when a mixture Pt/α-Al 2of Otemperature 3the catalyst (at point B) reduces the points sensor response AuPtPd/SnO 2 catalyst to H2the willtemperature be inhibited.difference The designbetween and composition of A of H and CO is introduced into the calorimetric-TGS device, the sensor response of the AuPtPd/SnO 2 single and B. Asand a result, when a mixture of H2 andofCO introduced2 into the calorimetric-TGS the 2 double catalyst type structures the is AuPtPd/SnO and 0.3–30 wt% Pt/α-Al2O3device, catalysts catalyst toresponse H inhibited. The design and composition of single and double catalyst type sensor the AuPtPd/SnO 2 catalyst to H2 1. will be inhibited. The design and composition of 2 willofbe of the micro-TGSs are given in Figure 3 and Table structures of the AuPtPd/SnO 0.3–30 wt% of the micro-TGSs given in single and double catalyst type structures of thePt/α-Al AuPtPd/SnO 2 and 0.3–30 wt% Pt/α-Al2Oare 3 catalysts 2 and 2 O3 catalysts Figure 3 and Table 1.are given in Figure 3 and Table 1. of the micro-TGSs
(a)
(a)
(b)
Pt line Heater
Pt line
A: Hot side (AuPtPd/SnO2)
CO or H2 + O2
(b)
B: Cold side Thermoelectric (Pt/a-Al2O3)
SiGe line
2H2+O2
+ O2
2H2O
Oxidation
Oxidation
CO or H2 A: Hot side
B: Cold side Heater (Pt/a-Al2O3)
A: Hot side A: Hot side Si membrane (AuPtPd/SnO2)
CO2 or H2O
CO2 or H2O
2H2+O2
Oxidation
Oxidation Thermoelectric
1 wt%Au, 1 wt%Pt, 1 wt%Pd /SnO2 catalyst
B: Cold side
2H2O
B: Cold side
30-0.3 wt% Pt /a-Al2O3 catalysts
Thermoelectric
Si membrane
Thermoelectric 1 wt%Au, 1 wt%Pt, 1 wt%Pd 30-0.3 wt% Pt SiGe line /a-Al2O3 catalysts Figure 2. Optical images of the AuPtPd/SnO2 and/SnO Pt/Al 2O3 catalysts on the Si membrane of the 2 catalyst micro-TGS (a), and a side view of the double catalyst structure (b).
Figure Optical imagesofof the AuPtPd/SnO2 and and Pt/Al2OO 3 catalysts on the Si membrane of the 3 Pt/Al Figure 2. 2. Optical images the AuPtPd/SnO 2 2 3 catalysts on the Si membrane of the micro-TGS (a), and a side view of the double catalyst structure (b). micro-TGS (a), and a side view of the double catalyst structure (b).
3
31689
Sensors 2015, 15, 31687–31698
Sensors 2015, 15, page–page
0.3 wt%
(b)
(a)
A
A
B 3 wt%
(c)
A
B 30 wt%
(d)
A
B
B
Figure 3. Double catalyst AuPtPd/SnO and 0.3–30 wt% Pt/α-Al Figure 3. Double catalyststructures structures with with AuPtPd/SnO 2 (A: HotHot side)side) and 0.3–30 wt% Pt/α-Al 2O3 (B: 2 O3 2 (A: (B: Cold themicro-TGS: micro-TGS:(a)(a)No No catalyst; Pt catalyst; (c) 3 Pt wt% Pt catalyst, Coldside) side)catalysts catalysts for the PtPt catalyst; (b)(b) 0.3 0.3 wt%wt% Pt catalyst; (c) 3 wt% catalyst, and (d) 30 wt% Pt catalyst. and (d) 30 wt% Pt catalyst. Table 1. Design of AuPtPd/SnO2 and and Pt/α-Al 2O3 double catalyst structures. Table 1. Design of AuPtPd/SnO Pt/α-Al 2 2 O3 double catalyst structures. Pt content in Pt/α-Al2O3
Notation
Design
0Pt
AuPtPd/SnO2 + no catalyst
Design
Notation 0Pt 0.3Pt 0.3Pt 3Pt 3Pt 30Pt 30Pt
Pt content in Pt/α-Al2 O3 wt% wt%
AuPtPd/SnO2 + no catalyst AuPtPd/SnO 2 + Pt/α-Al2O3 AuPtPd/SnO 2 + Pt/α-Al2 O3 AuPtPd/SnO AuPtPd/SnO2 + Pt/α-Al 2O 3 2 + Pt/α-Al 2 O3 AuPtPd/SnO2 + Pt/α-Al2 O3 AuPtPd/SnO2 + Pt/α-Al2O3
0
0
0.3 0.3 3 3 30
30
2.2. Catalyst Powder Characterization 2.2. Catalyst Powder Characterization morphology thecatalyst catalystpowder powder was electron microscopy The The morphology of of the was observed observedbybytransmission transmission electron microscopy (TEM; Tecnai Osiris, FEI, OR, USA) with high-angle annular dark field scanning transmission (TEM; Tecnai Osiris, FEI, OR, USA) with high-angle annular dark field scanning transmission electron electron microscopy (STEM-HAADF), at an accelerating 200Elemental kV. Elemental mapping microscopy (STEM-HAADF), at an accelerating voltagevoltage of 200ofkV. mapping of of Au, Pt, Au, Pt, Pd, and Sn was carried out using STEM-energy dispersive X-ray spectroscopy (STEM-EDS). Pd, and Sn was carried out using STEM-energy dispersive X-ray spectroscopy (STEM-EDS). CO CO adsorption and oxidation of the catalyst powder were investigated by diffuse reflectance Fourier adsorption and oxidation of the catalyst powder were investigated by diffuse reflectance Fourier transform Infrared spectroscopy (DRIFT) analysis (Nexus 470 FTIR, Nicolet, Waltham, MA, USA), at transform spectroscopy (DRIFT) analysis 470 FTIR, Nicolet, Waltham, MA, at 25, 50, Infrared 100, 200, and 300 °C. As measurement gases,(Nexus 10,000 ppm CO in dry air and 10,000 ppm COUSA), in ˝ 25, 50, and As of measurement gases, 10,000 in dryin airabsorbance and 10,000 ppm CO Ar100, were200, used at 300 a flowC.rate 50 cm3/min. DRIFT spectra ppm were CO measured mode 3 in Arwithout were used at a flow 50orcm switching carrierrate gas of (Air Ar)./min. DRIFT spectra were measured in absorbance mode
without switching carrier gas (Air or Ar). 2.3. Investigation of Gas Sensing Properties
2.3. Investigation of Gas Sensing Properties
Measurement system of sensing properties of micro-TGS is shown in Figure 4. The sensor response of micro-TGS 2, CO, and CH4 wasofinvestigated a gas flow chamber a Measurement systemfor of H sensing properties micro-TGSusing is shown in Figure 4. with The sensor volume of 60 cm3. The to the micro-TGS device wasaadjusted tochamber control the heater response of micro-TGS for voltage H2 , CO,applied and CH was investigated using gas flow with a volume 4 (50–250 °C) and was monitored using an IR camera (LAIRD-270A, Nikon Co., Tokyo, of 60temperatures cm3 . The voltage applied to the micro-TGS device was adjusted to control the heater temperatures Japan). To test the micro-TGS, pure dry air was allowed to flow into the chamber together with the (50–250 ˝ C) and was monitored using an IR camera (LAIRD-270A, Nikon Co., Tokyo, Japan). To test test gases (in dry air) in the following order: Dry air, test gas, and dry air. Each step lasted 90 s with a the micro-TGS, pure dry air was allowed to flow into the chamber together with the test gases (in dry gas flow rate of 200 cm3/min. The voltage across the thermoelectric thin film of the micro-TGS varied air) in the following air, test gas, and dry air. Each stepSeebeck lasted principle. 90 s withThe a gas flow rate by the combustionorder: heat ofDry the catalysts, a phenomenon based on the voltage 3 /min. The voltage across the thermoelectric thin film of the micro-TGS varied by the of 200 cm signal (Vs) was corrected using a digital multimeter (K2700, TFF Co. Keithley Inst., Tokyo, Japan) combustion of the catalysts, a phenomenon based onwith the 10,000 Seebeck principle. voltage followingheat pretreatment of the double catalyst micro-TGS ppm H2 in dryThe air at 200 °C.signal The response of using Vs (ΔV) is defined by Equation (1): TFF Co. Keithley Inst., Tokyo, Japan) following (Vs) was corrected a digital multimeter (K2700,
pretreatment of the double catalyst micro-TGS with 10,000 ppm H2 in dry air at 200 ˝ C. The response 4 of Vs (∆V) is defined by Equation (1): ∆V “ Vsp170´180q ´ Vsp80´90q 31690
(1)
Sensors 2015, 15, page–page
ΔV = Vs(170–180) – Vs(80–90)
Sensors 2015, 15, page–page
(1)
Sensors 15, 31687–31698 where2015, Vs(170–180) and Vs(80–90)
are the averages of Vs over 170–180 s (in test gas) and 80–90 s (in dry air), respectively. The sensor response was investigated the voltage across the micro-TGS ΔV = Vs(170–180) by – Vsmeasuring (80–90) (1) at 1000 ppm CO, H2, and CH4 at 50–250 °C and 10–10,000 ppm CO or H2 at 200 °C. The stability of the where Vs(170–180) andVs Vsp(80–90) thethe averages of Vsofover 170–180 s (in test gas) andgas) 80–90 s (in dry air), where Vsp170–180 averages Vs over 170–180 s (in test and 80–90 s (in dry q and 80–90are q are sensor responses was determined by subjecting the micro-TGS to a gas sensitivity test at 200 °C, for respectively. The was investigated by by measuring the voltage across the micro-TGS air), respectively. Thesensor sensorresponse response was investigated measuring the voltage across the micro-TGS 1, 2, at and 9 days.CO, H2, and CH4 at 50–250˝°C and 10–10,000 ppm CO or H2 at 200 °C. The 1000 of theof the at 1000 ppmppm CO, H2 , and CH4 at 50–250 C and 10–10,000 ppm CO or H2 at 200 ˝ C. stability The stability sensor responses was determined by subjecting the micro-TGS to a gas sensitivity test at 200 °C,˝for sensor responses was determined by subjecting the micro-TGS to a gas sensitivity test at 200 C, for 1, 1, 2, and 9 days. 2, and 9 days. Control of the gas flow rate and switching of gas type Control of the gas flowSignal rate and switching of gas type from multimeter
PC
Signal fromsensor multimeter Image Signal from
Digital multimeter Digital multimeter
Gas flow controller
Gas tank
CCD camera
CCD camera
Gas flow controller
Gas tank
PC
Image
Signal from sensor
Gas flow Micro-TGS
Gas flow
Micro-TGS
Measuring
Waste gas Waste gas
Measuring chamber chamber
Stabilized power
Stabilized supply power supply
Gas supply Gas supply
Figure 4. of gas sensing properties of micro-TGS. micro-TGS. Figure 4. Measurement Measurement system of gas gassensing sensingproperties properties of Figure 4. Measurementsystem system of of micro-TGS.
3. Results and Discussion 3. Results and Discussion 3. Results and Discussion 3.1. Catalyst Microstructure 3.1. Catalyst Microstructure Figure 5 shows the TEM images of the AuPtPd/SnO2 catalyst powder. Both well-dispersed Figure showsthe theTEM TEM images of the AuPtPd/SnO 2 catalyst powder. well-dispersed 55shows images of the AuPtPd/SnO powder. Both Both well-dispersed SnO2 2 catalyst SnO2 particles (20–50 nm) (Figure 5a), and small metal particles (3 nm) supported on SnO2 particles SnO2 particles (20–50 nm) (Figure 5a),small and small metal particles (3 supported nm) supported on 2SnO 2 particles particles (20–50 nm) (Figure 5a), and metal particles (3 nm) on SnO particles can clearly be seen (Figure 5b). It should be noted here that the high combustion performance of the can can clearly be seen (Figure 5b). It should notedhere herethat that The thehigh high combustion performance of the clearly be seen (Figure It should bebe noted performance catalyst is affected by5b). its porous structure and particle size.the graincombustion size of the aggregated metalof the catalyst is affected by its porous structure and particle size. The grain size of the aggregated metal catalyst is affected by itsnanometers, porous structure particle size. The grain sizeresulting of the aggregated particles was several similarand to the starting colloidal particle, in relativelymetal particles was severalnanometers, nanometers, similar the starting colloidal particle, resulting inAu) relatively particles several similar to to the starting colloidal particle, resulting relatively goodwas dispersion. However, to investigate whether each metal element (i.e., Pt, Pd,inand is good good dispersion. However, to investigate whether each metal element (i.e., Pt, Pd, and is alloyed or not, the catalyst was analyzed by STEM and elemental mapping. dispersion. However, to investigate whether each metal element (i.e., Pt, Pd, and Au) is alloyed Au) or not,
alloyed or not, catalystby was analyzed by STEMmapping. and elemental mapping. the catalyst wasthe analyzed STEM and elemental
Figure 5. Transmission electron microscopy (TEM) images of the AuPtPd/SnO2 catalyst powder: (a) low magnification (17,500) and (b) high magnification (88,000).
5 5. Transmission electron microscopy (TEM) (TEM) images images of of the the AuPtPd/SnO AuPtPd/SnO22 catalyst Figure 5. catalyst powder: powder: (17,500) and and (b) (b) high high magnification magnification (88,000). (88,000). (a) low magnification (17,500)
5
31691
Sensors 2015, 15, 31687–31698 Sensors 2015, 15, page–page
Figure 6 shows the High-angle annular dark field scanning transmission electron microscopy Figure 6images shows the annular dark field scanning transmission electron microscopy (STEM-HAADF) andHigh-angle STEM-energy dispersive X-ray spectroscopy (STEM-EDS) mapping of the (STEM-HAADF) images and STEM-energy dispersive X-ray spectroscopy (STEM-EDS) mapping of Au, Pt, Pd, and Sn distributions in the catalyst powder. Low magnification images (Figure 6a) indicate the Au, Pt, Pd, and Sn distributions in the catalyst powder. Low magnification images (Figure 6a) that the noble metals were dispersed over the SnO2 particles. However, in the high magnification indicate that the noble metals were dispersed over the SnO2 particles. However, in the high images (Figure 6b), only the Pt and Pd nanoparticles were deposited and dispersed on SnO2 particles, magnification images (Figure 6b), only the Pt and Pd nanoparticles were deposited and dispersed on while the Au nanoparticles aggregated as 10 nm particles, overlapping with Pt and Pd, as indicated by SnO2 particles, while the Au nanoparticles aggregated as 10 nm particles, overlapping with Pt and arrows overlap 6b. image of Figure 6b. Pd,in asthe indicated byimage arrowsof inFigure the overlap
Figure 6. High-angle annular dark field scanning transmission electron microscopy (STEM-HAADF)
Figure 6. High-angle annular dark field scanning transmission electron microscopy (STEM-HAADF) and STEM-energy dispersive X-ray spectroscopy (STEM-EDS) images of the AuPtPd/SnO2 catalyst and STEM-energy dispersive image X-ray and spectroscopy images of the AuPtPd/SnO catalyst powder. (a) STEM-HAADF EDS maps (STEM-EDS) of Au, Pt, Pd, Sn, and Overlap (Au-Pt-Pd) at2low powder. (a) STEM-HAADF image and EDS maps of Au, Pt, Pd, Sn, and Overlap (Au-Pt-Pd) at low magnification (110,000), and (b) STEM-HAADF image and EDS maps of Au, Pt, Pd, Sn, and Overlap magnification (110,000), and (b) STEM-HAADF image and EDS maps of Au, Pt, Pd, Sn, and Overlap at at high magnification (630,000). high magnification (630,000). 3.2. DRIFT Characterization of the Catalyst
3.2. DRIFT Characterization the Catalyst Figure 7 shows the of 1000–4000 cm−1 region of the diffuse reflectance Fourier transform Infrared spectroscopy (DRIFT) spectra of the AuPtPd/SnO2 catalyst at 25–300 °C in 10,000 ppm CO in both dry
Figure 7 shows the 1000–4000 cm´1 region of the diffuse reflectance Fourier transform Infrared air (Figure 7a) and in Ar (Figure 7b). As can be seen in Figure 7a, at lower temperatures (25–100 °C), spectroscopy (DRIFT) spectra of the AuPtPd/SnO at 25–300 ˝ C in 10,000 ppm CO in both dry 2 catalyst bands at 1082, 1221, 1311, and 1551 cm−1 were observed, corresponding to vibrations of the ˝ air (Figure 7a) species and in Ar (Figure 7b). As can2be seen in Figure lower temperatures (25–100 carbonate on noble metals or SnO [15,20]. The bands7a, at at 1834, 2078, 2116, and 2175 cm−1 C), ´ 1 bandscould at 1082, 1221, 1311, and cm the were observed, to broad vibrations of 1834 the carbonate be assigned to CO. In1551 particular, sharp peak at corresponding 2078 cm−1 and the peak at cm−1 ´ 1 species on nobleto metals or SnO bands at 1834, 2116, and 2175respectively cm could be assigned correspond the linear and bridgingThe bonds of the CO-Pt2078, and CO-Pd bands, [21]. The 2 [15,20]. 1 and the 1 correspond to peak corresponding to COat adsorption decreased with increasing the1834 catalyst to CO. In intensity particular, the sharp peak 2078 cm´ broad peak at cm´temperature. At catalyst temperatures of of 200the andCO-Pt 300 °C, bands at 2314 andrespectively 2359 cm−1, corresponding to CO 2, the linear and bridging bonds and CO-Pd bands, [21]. The peak intensity were observed, confirming the oxidation of CO by the AuPtPd/SnO 2 catalyst. For the experiments corresponding to CO adsorption decreased with increasing the catalyst temperature. At catalyst employing 10,000 ppm CO in Ar (Figure 7b), the peaks corresponding to the carbonate species (at temperatures of 200 and 300 ˝ C, bands at 2314 and 2359 cm´1 , corresponding to CO2 , were observed, 1097, 1225, 1313, 1549, 2079, and 2172 cm−1) were detected at 25–200 °C. However, in contrast to the confirming the oxidation of CO by the AuPtPd/SnO2 catalyst. For the experiments employing experiments employing 10,000 ppm CO in dry air (Figure 7a), peaks corresponding to CO2 were 10,000 ppm CO in Ar (Figure 7b), the peaks corresponding to the carbonate species (at 1097, 1225, not detected. 1313, 1549, 2079, and 2172 cm´1 ) were detected at 25–200 ˝ C. However, in contrast to the experiments employing 10,000 ppm CO in dry air (Figure 7a), peaks corresponding to CO2 were not detected.
6
31692
Sensors 2015, 15, 31687–31698
Sensors 2015, 15, page–page
100 ºC 50 ºC
1549
1313 1225 1097
2079
2172
2357
2550
300 ºC 1834
(a.u.) Absorbance/ a.u. Absorbance
1221 1082
1834
200 ºC
3691
1573
1446 1311 1171 1128
2359 2314 2175 2116 2078
300 ºC 1551
Absorbance / a.u. Absorbance (a.u.)
2553
(b) 3701
(a)
200 ºC 100 ºC 50 ºC 25 ºC
25 ºC
4000
3000 2000 -1) Wavenumber /(cm cm -1 Wavenumber
4000
1000
3000 2000 Wavenumber Wavenumber /(cm cm -1-1)
1000
FigureFigure 7. Diffuse reflectance Infrared spectroscopy spectroscopy (DRIFT) spectra of the 7. Diffuse reflectanceFourier Fourier transform transform Infrared (DRIFT) spectra of the ˝ C, in (a) 10,000 ppm CO in dry air, and (b) 10,000 ppm CO AuPtPd/SnO catalyst powder at 25–300 2 AuPtPd/SnO 2 catalyst powder at 25–300 °C, in (a) 10,000 ppm CO in dry air, and (b) 10,000 ppm CO in Ar.in Ar.
The DRIFT spectra (Figure 7) clearly showed a peak corresponding to CO adsorption on the Pt
The DRIFT spectra (Figure 7) clearly showed a peak corresponding to CO adsorption on the and Pd particles at 2078 cm−1, but no sharp peak corresponding to CO adsorption on the metallic Au Pt and Pd particles at 2078 cm´−11 , but no sharp peak corresponding to CO adsorption on the metallic site was detected at 2112 cm as shown in previous researches [15,22]. This indicated that less CO ´1 as shown in previous researches [15,22]. This indicated that less Au site was detected at 2112 cmAu oxidation took place on the particles compared to the Pt and Pd particles. As shown in the CO oxidation took place on the6),Au to the Ptthus andincreasing Pd particles. shown in the STEM-EDS images (Figure Auparticles particles compared formed aggregates, theirAs particle size STEM-EDS images (Figure 6), Au particles formed aggregates, thus increasing their particle (≥10 nm). This can be explained by previous studies showing that the oxidation reaction of the Au size catalyst slowed with increasingby particle diameter, with a particlethat sizethe of 30Pt. these results, 0Pt > 0.3Pt ~ 3Pt > 30Pt. Considering these results, an optimal Pt content of 0.3Pt2O~3 3Pt > 30Pt. Considering these results, Pt content of
