NO2 Sensing Capability of Pt–Au–SnO2 Composite Nanoceramics at Room Temperature

Composite ceramics of metal oxides and noble metals have received much attention for sensing reducing gases at room temperature. Presently, composite ceramics of SnO2 and noble metals have been prepared and investigated for sensing oxidizing NO2 at room temperature. While dramatic increases in resistance were observed for both 1 wt% Pt–SnO2 and 5 wt% Au–SnO2 composite nanoceramics after being exposed to NO2 at room temperature, the largest increase in resistance was observed for 1 wt% Pt–5 wt% –Au–SnO2 composite nanoceramics among the three composites. The response to 0.5 ppm NO2-–20% O2–N2 was as high as 875 at room temperature, with a response time of 2566 s and a recovery time of 450 s in the air of 50% relative humidity (RH). Further investigation revealed that water molecules in the air are essential for recovering the resistance of Pt–Au–SnO2 composite nanoceramics. A room temperature NO2-sensing mechanism has been established, in which NO2 molecules are catalyzed by Pt–Au to be chemisorbed on SnO2 at room temperature, and desorbed from SnO2 by the attraction of water molecules in the air. These results suggest that composite ceramics of metal oxides and noble metals should be promising for room temperature sensing, not only reducing gases, but also oxidizing gases.


Introduction
Due to more and more fuel combustion in places such as thermal power plants and automobiles, NO 2 in the air has been increasing quickly in recent years [1,2]. NO 2 is not only directly harmful to human health, but also results in soil contamination through the formation of acid rain [3][4][5][6]. With the advantages of good sensitivity and selectivity, electrochemical gas sensors have been most widely used for NO 2 detection [7,8]. However, their relatively short service life [8], susceptibility to environmental interference, and poor stability, have caused much inconvenience for NO 2 detection. Other NO 2 gas sensors, with good stability, long service life, and low price, are highly expected.
There have been extensive investigations devoted to developing NO 2 gas sensors based on metal oxides [9][10][11][12][13][14], and many of them have been focused on developing roomtemperature metal oxide NO 2 gas sensors. Room temperature operation is not only important for low power consumption, but also for miniaturization. Several research groups have made some impressive progresses by adopting nanostructured metal oxides. For example, Zhang et al. have prepared SnO 2 -ZnO with a layered nanostructure, which showed responses to ppb level NO 2 at 150 • C [15]. Bang  which achieved room-temperature NO 2 detection by visible light irradiation [18]. Pham et al. have fabricated MoS 2 single-layer films, which could detect NO 2 at room temperature with red-light irradiation [19]. Although the introduction of light irradiation has brought the operating temperature down to room temperature, it will cause some inconvenience in itself. It has to be pointed out that these low-dimensional-nanostructured materials that have been investigated are of poor mechanical strength and are relatively complicated in composition and structure, which will be unfavorable for practical applications. Many more explorations are highly desirable to develop relatively simple and robust room-temperature NO 2 sensors based on metal oxides.
As a matter of fact, to develop room-temperature gas sensors based on metal oxides with a relatively high mechanical strength, two different strategies have already emerged. In the first strategy, a porous nanosolid (PNS) is prepared from metal oxide nanoparticles through a solvothermal hot press (SHP) [20]. PNS is considered an intermediate state between nanoparticles and nanoceramics [21,22], and with both high reactivity of nanoparticles and strength of nanoceramics: some impressive room temperature metaloxide gas sensors have been fabricated from PNSs [23,24]. In the second strategy, the catalytic effect of noble metals (e.g., Pt and Pd) is utilized to achieve room-temperature gas sensing. Highly remarkable room-temperature gas sensing capabilities have been observed in many composite ceramics of noble metals and metal-oxide semiconductors (e.g., TiO 2 , WO 3 , Nb 2 O 5 , SnO 2 and ZnO), which were prepared through traditional pressing and sintering [25][26][27][28][29][30]. It is worthy to note that for composites of Pt with micron-sized WO 3 and SnO 2 agglomerate powder [26,27], surprisingly strong room-temperature responses to hydrogen and an extraordinarily high moisture resistance have been observed, which is especially important for practical room-temperature gas sensing applications. Obviously, this strategy is highly appealing for developing room-temperature metal-oxide gas sensors, with promising practical application potentials. However, it has to be pointed out that bulk ceramics prepared through this strategy can only sense reducing gases of H 2 and CO at room temperature at present, in which noble metals, such as Pt and Pd, promote H 2 and CO to react with the oxygen chemisorbed on the metal oxides, and even be chemisorbed on the metal oxides at room temperature [31][32][33][34][35]. There have been no reports on preparing bulk materials capable of sensing oxidizing gases, including NO 2 , at room temperature through this strategy up to date.
In a previous investigation, the resistance of Pt-SnO 2 nanoparticles was found to increase dramatically with increasing Pt content, which clearly indicates that Pt can promote the chemisorption of oxygen molecules on SnO 2 at room temperature [36]. In another investigation, Pd was also found to be able to promote oxygen chemisorption on SnO 2 in Pd-SnO 2 nanoparticles at room temperature through XPS analyses [29]. It is well known that O 2 is a typically oxidizing gas. These facts suggest that noble metals, such as Pt and Pd, may also be able to promote some oxidizing gases to be chemisorbed on some metal oxides at room temperature, and in turn, their composites, with these metal oxides, should be able to show responses to the oxidizing gases at room temperature. Thus, this indicates that bulk composites capable of sensing oxidizing gases at room temperature should be possibly obtained through the second strategy. Presently, we have adopted this strategy to prepare Pt-SnO 2 composite nanoceramics through pressing and sintering, which were indeed found to show strong responses to NO 2 at room temperature. More interestingly, the roomtemperature NO 2 sensing characteristics were further dramatically improved through the introduction of Au to the composites, and a rather remarkable room-temperature NO 2 sensing capability has been observed for Pt-Au-SnO 2 composite nanoceramics. Some further studies have been conducted, which show that water molecules in the air play a vital role for those samples to recover their resistance in the air after being exposed to NO 2 . It is proposed that NO 2 molecules are catalyzed by Pt-Au to be chemisorbed on SnO 2 at room temperature, and are removed from SnO 2 by the attraction of water molecules in the air. These results clearly demonstrate that metal-oxide bulk materials capable of sensing oxidizing gases at room temperature can be prepared through pressing and sintering. However, more investigation regarding this is highly desirable. Figure 1 shows the X-ray diffraction (XRD) patterns obtained for three kinds of nanoceramics, of which the nominal compositions/sintering temperatures are 1 wt% Pt-SnO 2 /950 • C, 5 wt% Au-SnO 2 /950 • C and 1 wt% Pt-5 wt% Au-SnO 2 /950 • C, respectively. As shown in Figure 1a, for 1 wt% Pt-SnO 2 , three diffraction peaks can be clearly identified as (111), (200) and (220) planes of cubic Pt, according to JCPDS 04-0802. All other diffraction peaks can be identified as planes of rutile SnO 2 , according to CPDS 78-1063. Obviously, this sample was a composite of the cubic Pt and rutile SnO 2 . Figure 1b illustrates the XRD diffraction pattern of 5 wt% Au-SnO 2 . Similarly, all the diffraction peaks can be identified as planes of rutile SnO 2 and cubic Au (JCPDS 65-2870) respectively, indicating a composite of Au and SnO 2 . Figure 1c presents the XRD pattern of 1 wt% Pt-5 wt% Au-SnO 2 . Beside the peaks from the rutile SnO 2 , some peaks from both the cubic Au and Pt can be observed, which indicates that Pt and Au existed as separate phases in this sample. Therefore, this sample of 1 wt% Pt-5 wt% Au-SnO 2 was a composite of the cubic Pt, cubic Au, and rutile SnO 2 . As a matter of fact, both Pt and Au are highly stable noble metals, and it is reasonable that they can form composites with SnO 2 through high-temperature sintering.

Phase and Morphological Investigations
Molecules 2023, 28, x FOR PEER REVIEW 3 of 11 SnO2 at room temperature, and are removed from SnO2 by the attraction of water molecules in the air. These results clearly demonstrate that metal-oxide bulk materials capable of sensing oxidizing gases at room temperature can be prepared through pressing and sintering. However, more investigation regarding this is highly desirable. Figure 1 shows the X-ray diffraction (XRD) patterns obtained for three kinds of nanoceramics, of which the nominal compositions/sintering temperatures are 1 wt% Pt-SnO2/950 °C, 5 wt% Au-SnO2/950 °C and 1 wt% Pt-5 wt% Au-SnO2/950 °C, respectively. As shown in Figure 1a, for 1 wt% Pt-SnO2, three diffraction peaks can be clearly identified as (111), (200) and (220) planes of cubic Pt, according to JCPDS 04-0802. All other diffraction peaks can be identified as planes of rutile SnO2, according to CPDS 78-1063. Obviously, this sample was a composite of the cubic Pt and rutile SnO2. Figure 1b illustrates the XRD diffraction pattern of 5 wt% Au-SnO2. Similarly, all the diffraction peaks can be identified as planes of rutile SnO2 and cubic Au (JCPDS 65-2870) respectively, indicating a composite of Au and SnO2. Figure 1c presents the XRD pattern of 1 wt% Pt-5 wt% Au-SnO2. Beside the peaks from the rutile SnO2, some peaks from both the cubic Au and Pt can be observed, which indicates that Pt and Au existed as separate phases in this sample. Therefore, this sample of 1 wt% Pt-5 wt% Au-SnO2 was a composite of the cubic Pt, cubic Au, and rutile SnO2. As a matter of fact, both Pt and Au are highly stable noble metals, and it is reasonable that they can form composites with SnO2 through high-temperature sintering. As pointed out in some previous papers, the SnO2 nanoparticles used in this study exhibit a very unique sintering behavior, and pellets prepared from them showed no sintering shrinkage, even after being sintered at 1200 °C [29]. Accordingly, all the samples prepared in this study showed no noticeable sintering shrinkage either. Figure 2a shows an SEM micrograph obtained for a fractured surface of a sample of 1 wt% Pt-5 wt% Au-SnO2 nanoceramics sintered at 950 °C for 2 h in the air. Firstly, some nanopores can be clearly observed in the micrograph, which should be helpful for gas sensing. It can be observed that most grains are about 70 nm in diameter, while a few much larger grains, around 300 nm in diameter, can also be observed. According to the EDS analysis shown As pointed out in some previous papers, the SnO 2 nanoparticles used in this study exhibit a very unique sintering behavior, and pellets prepared from them showed no sintering shrinkage, even after being sintered at 1200 • C [29]. Accordingly, all the samples prepared in this study showed no noticeable sintering shrinkage either. Figure 2a shows an SEM micrograph obtained for a fractured surface of a sample of 1 wt% Pt-5 wt% Au-SnO 2 nanoceramics sintered at 950 • C for 2 h in the air. Firstly, some nanopores can be clearly observed in the micrograph, which should be helpful for gas sensing. It can be observed that most grains are about 70 nm in diameter, while a few much larger grains, around 300 nm in diameter, can also be observed. According to the EDS analysis shown in Figure 2b, those smaller grains should be SnO 2 nanograins, and they must have experienced no obvious grain growth in the sintering, as they were quite similar to SnO 2 nanoparticles in size. The much larger grains were Pt and Au grains, respectively. The Au powder and Pt powder were marked as <500 nm and <1 µm, respectively. Thus, these large grains must have come from their starting materials. According to two very recent papers, large Pt grains are important for room-temperature CO-sensitive and H 2 -sensitive Pt-SnO 2 composite nanoceramics to achieve a high, long-term stability [37,38]. As a matter of fact, these two commercial Au and Pt powders, with relatively large particles, had been intentionally chosen as starting materials in this study.

Phase and Morphological Investigations
Molecules 2023, 28, x FOR PEER REVIEW 4 of 1 in Figure 2b, those smaller grains should be SnO2 nanograins, and they must have experi enced no obvious grain growth in the sintering, as they were quite similar to SnO2 nano particles in size. The much larger grains were Pt and Au grains, respectively. The Au pow der and Pt powder were marked as <500 nm and <1 μm, respectively. Thus, these larg grains must have come from their starting materials. According to two very recent papers large Pt grains are important for room-temperature CO-sensitive and H2-sensitive Pt SnO2 composite nanoceramics to achieve a high, long-term stability [37,38]. As a matter o fact, these two commercial Au and Pt powders, with relatively large particles, had bee intentionally chosen as starting materials in this study.

Room-Temperature NO2-Sensing Measurement
Pt-SnO2 composite nanoceramics have been found to show strong responses to th reducing gases of H2 and CO at room temperature, which is characterized by a dramati decrease in their resistance upon being exposed to the gases [29,36]. According to ou knowledge, however, there have been no reports on the room-temperature responses o the Pt-SnO2 composite nanoceramics to any oxidizing gases up to date. It is, thus, ver surprising to see that the Pt-SnO2 composite nanoceramics prepared in this study exhib ited an extraordinarily strong response to 10 ppm NO2-20% O2-N2 at room temperature as shown in Figure 3. For the sensing of oxidizing gases, the response S is usually define as S = Rg/Ra, where Rg and Ra represent the resistance of the sensor in the target gas and i the air, respectively, and response (recovery) time is defined as the time taken by the sen sor to reach 90% of the total resistance change after the introduction (discontinuation) o the gas for testing [15]. According to this definition, this sample of 1 wt% Pt-SnO2 had room-temperature response of 923 with a response time of 614 s and a recovery time o 1350 s to 10 ppm NO2-20% O2-N2. Such a room temperature response to NO2 is highl outstanding when compared with those of the newly reported metal oxides in the litera ture.
For reference, other kinds of composites had also been prepared and investigated fo room-temperature NO2 sensing. A very interesting result was observed for a sample of wt% Au-SnO2, sintered at 950 °C for 2 h in the air, as shown in Figure 3. Firstly, it can b seen that this sample had a very low resistance in the air, which indicates that Au is muc

Room-Temperature NO 2 -Sensing Measurement
Pt-SnO 2 composite nanoceramics have been found to show strong responses to the reducing gases of H 2 and CO at room temperature, which is characterized by a dramatic decrease in their resistance upon being exposed to the gases [29,36]. According to our knowledge, however, there have been no reports on the room-temperature responses of the Pt-SnO 2 composite nanoceramics to any oxidizing gases up to date. It is, thus, very surprising to see that the Pt-SnO 2 composite nanoceramics prepared in this study exhibited an extraordinarily strong response to 10 ppm NO 2 -20% O 2 -N 2 at room temperature, as shown in Figure 3. For the sensing of oxidizing gases, the response S is usually defined as S = R g /R a , where R g and R a represent the resistance of the sensor in the target gas and in the air, respectively, and response (recovery) time is defined as the time taken by the sensor to reach 90% of the total resistance change after the introduction (discontinuation) of the gas for testing [15]. According to this definition, this sample of 1 wt% Pt-SnO 2 had a room-temperature response of 923 with a response time of 614 s and a recovery time of 1350 s to 10 ppm NO 2 -20% O 2 -N 2 . Such a room temperature response to NO 2 is highly outstanding when compared with those of the newly reported metal oxides in the literature.
For reference, other kinds of composites had also been prepared and investigated for room-temperature NO 2 sensing. A very interesting result was observed for a sample of 5 wt% Au-SnO 2 , sintered at 950 • C for 2 h in the air, as shown in Figure 3. Firstly, it can be seen that this sample had a very low resistance in the air, which indicates that Au is much less effective than Pt in promoting the chemisorption of O 2 molecules on SnO 2 at room temperature. Secondly, this sample also showed a strong response to 10 ppm NO 2 -20% O 2 -N 2 at room temperature, with a response of 132, a response time of 1279 s and a recovery of 672 s. It seems that Au has a stronger catalytic effect on NO 2 molecules than O 2 molecules, with regards to their chemisorption on SnO 2 at room temperature.
cates that the room temperature chemisorption of O2 molecules on SnO2 in the presence of both Pt and Au is much different from that in the presence of Pt alone. Secondly, this sample showed an extraordinarily strong response to NO2 at room temperature, with a response of 6031, a response time of 591 s, and a recovery time of 430 s to 10 ppm NO2-20% O2-N2. Due to its rather small resistance in the air, the signal was quite stable in NO2, even with such a strong response. It is worthy to note that this sample had the strongest response to NO2 at room temperature among the three samples, which is actually very difficult to understand. Figure 3. Room-temperature response to 10 ppm NO2 in 20% O2-N2 and recovery in the air of 50% RH for three samples with compositions of 1 wt% Pt-SnO2, 5 wt% Au-SnO2 and 1 wt% Pt-5 wt% Au-SnO2, respectively.
The sample of 1 wt% Pt-5 wt% Au-SnO2 had been exposed to NO2 in a series of concentrations at room temperature and the results are shown in Figure 4. Its room-temperature response to 0.5 ppm NO2-20% O2-N2 was 875, with a response time of 2566 s, and a recovery time of 450 s. With increasing NO2 concentration, both the response and the response speed increased steadily. To 8 ppm NO2-20% O2-N2, the response was increased to 4300 with a much smaller response time of 924 s and a recovery time of 440 s. It is For the sample of 1 wt% Pt-SnO 2 , the response was very attractive, but its resistance to NO 2 was too high to ensure a stable measurement. The signal was a little unstable at the top of the curve. When regarding the sample of 5 wt% Au-SnO 2 , it had a much smaller resistance in NO 2 , but its response to NO 2 was also much weaker than that of the former sample. Clearly, these two samples had some rather complementary advantages and disadvantages. Therefore, we had prepared composites of 1 wt% Pt-5 wt% Au-SnO 2 , and the result was very surprising. As shown in Figure 3, impressive room-temperature NO 2 -sensing properties are observed for a sample of 1 wt% Pt-5 wt% Au-SnO 2 . First, this sample had a much lower resistance in the air than the sample of 1 wt% Pt-SnO 2 , which indicates that the room temperature chemisorption of O 2 molecules on SnO 2 in the presence of both Pt and Au is much different from that in the presence of Pt alone. Secondly, this sample showed an extraordinarily strong response to NO 2 at room temperature, with a response of 6031, a response time of 591 s, and a recovery time of 430 s to 10 ppm NO 2 -20% O 2 -N 2 . Due to its rather small resistance in the air, the signal was quite stable in NO 2 , even with such a strong response. It is worthy to note that this sample had the strongest response to NO 2 at room temperature among the three samples, which is actually very difficult to understand.
The sample of 1 wt% Pt-5 wt% Au-SnO 2 had been exposed to NO 2 in a series of concentrations at room temperature and the results are shown in Figure 4. Its roomtemperature response to 0.5 ppm NO 2 -20% O 2 -N 2 was 875, with a response time of 2566 s, and a recovery time of 450 s. With increasing NO 2 concentration, both the response and the response speed increased steadily. To 8 ppm NO 2 -20% O 2 -N 2 , the response was increased to 4300 with a much smaller response time of 924 s and a recovery time of 440 s. It is interesting to compare this sample with some representative NO 2 -sensitive nanomaterials newly reported in the literature, as shown in Table 1. It can be clearly seen that our sample had a much higher room-temperature response to a low concentration NO 2 than all these low-dimensional nanomaterials, and as a bulk material, our sample should also have a much higher mechanical strength. Obviously, these results indicate that Molecules 2023, 28, 1759 6 of 11 the composite nanoceramics of 1 wt% Pt-5 wt% Au-SnO 2 prepared in this study should be highly attractive for room-temperature NO 2 sensing.
interesting to compare this sample with some representative NO2-sensitive nanomaterials newly reported in the literature, as shown in Table 1. It can be clearly seen that our sample had a much higher room-temperature response to a low concentration NO2 than all these low-dimensional nanomaterials, and as a bulk material, our sample should also have a much higher mechanical strength. Obviously, these results indicate that the composite nanoceramics of 1 wt% Pt-5 wt% Au-SnO2 prepared in this study should be highly attractive for room-temperature NO2 sensing.

Mechanism Study on Room-Temperature NO2 Sensing Characteristics
As an oxidizing gas, NO2 increases the resistance of n-type semiconductors through its chemisorption, which is the same as what oxygen does. It is meaningful to identify the influence of NO2 from that of oxygen in NO2 sensing. For this purpose, we had exposed a sample of 1 wt% Pt-5 wt% Au-SnO2 to a series of specifically designated atmospheres at room temperature, as shown in Figure 5. When the atmosphere was first changed from air to 20% O2-N2, the resistance slowly increased with time, which actually indicates a drying effect of the flowing gas of 20% O2-N2 in the sample. Those oxygen molecules chemisorbed on SnO2 must be very stable, so there was no turning point when the atmosphere was changed from 20% O2-N2 to N2. Upon being exposed to 10 ppm NO2-N2, a steep increase was observed when the atmosphere was changed from N2 to 10 ppm NO2-N2. The resistance was finally increased by more than three orders of magnitude, which demonstrates a strong chemisorption of NO2 on SnO2. When the surrounding atmosphere

Mechanism Study on Room-Temperature NO 2 Sensing Characteristics
As an oxidizing gas, NO 2 increases the resistance of n-type semiconductors through its chemisorption, which is the same as what oxygen does. It is meaningful to identify the influence of NO 2 from that of oxygen in NO 2 sensing. For this purpose, we had exposed a sample of 1 wt% Pt-5 wt% Au-SnO 2 to a series of specifically designated atmospheres at room temperature, as shown in Figure 5. When the atmosphere was first changed from air to 20% O 2 -N 2 , the resistance slowly increased with time, which actually indicates a drying effect of the flowing gas of 20% O 2 -N 2 in the sample. Those oxygen molecules chemisorbed on SnO 2 must be very stable, so there was no turning point when the atmosphere was changed from 20% O 2 -N 2 to N 2 . Upon being exposed to 10 ppm NO 2 -N 2 , a steep increase was observed when the atmosphere was changed from N 2 to 10 ppm NO 2 -N 2 . The resistance was finally increased by more than three orders of magnitude, which demonstrates a strong chemisorption of NO 2 on SnO 2 . When the surrounding atmosphere was changed from 10 ppm NO 2 -N 2 to N 2 , the resistance only decreased very slowly with time, which further confirms a strong chemisorption of NO 2 on SnO 2 . No turning point appeared when the atmosphere was changed from N 2 to 20% O 2 -N 2 . It is worth mentioning that when the atmosphere was changed from 20% O 2 -N 2 to N 2 and from N 2 to 20% O 2 -N 2 , respectively, only the oxygen content was changed. The absence of the turning point for oxygen content change in the resistance versus the time curve indicates a stable oxygen chemisorption on SnO 2 . Considering that both NO 2 and O 2 are strongly chemisorbed on SnO 2 at room temperature, it was highly surprising to see that the resistance was sharply decreased when the surrounding atmosphere was changed from 20% O 2 -N 2 to the air of 50% RH, as shown in Figure 5. Given the difference between these two atmospheres, water molecules in the air had obviously played the vital role in this resistance decrease. appeared when the atmosphere was changed from N2 to 20% O2-N2. It is worth mentioning that when the atmosphere was changed from 20% O2-N2 to N2 and from N2 to 20% O2-N2, respectively, only the oxygen content was changed. The absence of the turning point for oxygen content change in the resistance versus the time curve indicates a stable oxygen chemisorption on SnO2. Considering that both NO2 and O2 are strongly chemisorbed on SnO2 at room temperature, it was highly surprising to see that the resistance was sharply decreased when the surrounding atmosphere was changed from 20% O2-N2 to the air of 50% RH, as shown in Figure 5. Given the difference between these two atmospheres, water molecules in the air had obviously played the vital role in this resistance decrease. Figure 5. Room-temperature resistance responses to a sequence of atmospheres: air, 20% O2-N2, N2, 10 ppm NO2-N2, N2, 20% O2-N2, and air for a sample of 1 wt% Pt-5 wt% Au-SnO2 sintered at 950 °C.
The chemisorption of NO2 on SnO2 at room temperature in 10 ppm NO2-N2 can be expressed as [43]: in which Pt-Au must have acted as the catalyst for the reaction, as shown in Figure 6a. As an oxidizing gas, NO2 molecules accept electrons from SnO2 when they are chemisorbed on SnO2. This explains the steep increase in resistance when the sample is exposed to 10 ppm NO2-N2. When the surrounding atmosphere is changed from 10 ppm NO2-N2 to N2, the chemisorption of NO2 is so stable that most NO2 molecules will not leave SnO2 in the surrounding N2, as shown in Figure 6b. In this way, only a slight decrease in resistance can be observed. When the surrounding atmosphere is changed from N2 to 20% O2-N2, no more O2 molecules can be chemisorbed on SnO2, as those O2 molecules chemisorbed on SnO2 earlier have remained there, as shown in Figure 6c. Thus, no resistance change can be observed for this atmosphere change. H2O and NO2 molecules are able to form NO2-H2O clusters through the hydrogen bonds between them [44]. When the ambient atmosphere is changed from 20% O2-N2 to the air of 50% relative humidity, NO2 molecules are desorbed from SnO2 by the attraction of H2O molecules, as shown in Figure 6d. This desorption of NO2 can be expressed as: The chemisorption of NO 2 on SnO 2 at room temperature in 10 ppm NO 2 -N 2 can be expressed as [43]: in which Pt-Au must have acted as the catalyst for the reaction, as shown in Figure 6a. As an oxidizing gas, NO 2 molecules accept electrons from SnO 2 when they are chemisorbed on SnO 2 . This explains the steep increase in resistance when the sample is exposed to 10 ppm NO 2 -N 2 . When the surrounding atmosphere is changed from 10 ppm NO 2 -N 2 to N 2 , the chemisorption of NO 2 is so stable that most NO 2 molecules will not leave SnO 2 in the surrounding N 2 , as shown in Figure 6b. In this way, only a slight decrease in resistance can be observed. When the surrounding atmosphere is changed from N 2 to 20% O 2 -N 2 , no more O 2 molecules can be chemisorbed on SnO 2 , as those O 2 molecules chemisorbed on SnO 2 earlier have remained there, as shown in Figure 6c. Thus, no resistance change can be observed for this atmosphere change. H 2 O and NO 2 molecules are able to form NO 2 -H 2 O clusters through the hydrogen bonds between them [44]. When the ambient atmosphere is changed from 20% O 2 -N 2 to the air of 50% relative humidity, NO 2 molecules are desorbed from SnO 2 by the attraction of H 2 O molecules, as shown in Figure 6d. This desorption of NO 2 can be expressed as: The electrons captured by the NO 2 molecules are returned to SnO 2 , so the resistance is greatly decreased when air with H 2 O molecules is introduced. This forms a sharp contrast with the resistance recovery in the air for n-type metal oxides after being exposed to reducing gases at room temperature, in which oxygen molecules in the air are chemisorbed on the metal oxides and the resistance is increased.
The electrons captured by the NO2 molecules are returned to SnO2, so the resistance is greatly decreased when air with H2O molecules is introduced. This forms a sharp contrast with the resistance recovery in the air for n-type metal oxides after being exposed to reducing gases at room temperature, in which oxygen molecules in the air are chemisorbed on the metal oxides and the resistance is increased.

Material Preparation
SnO2 nanoparticles (70 nm, 99.99%), a commercial Au powder (<500 nm, 99.9%), and a commercial Pt powder (<1 μm, 99.9%) from Aladdin, Shanghai, China, were used as the starting materials. According to designated ratios, these particles were mixed in deionized water and magnetically stirred. For every suspension, magnetic stirring was performed

Material Preparation
SnO 2 nanoparticles (70 nm, 99.99%), a commercial Au powder (<500 nm, 99.9%), and a commercial Pt powder (<1 µm, 99.9%) from Aladdin, Shanghai, China, were used as the starting materials. According to designated ratios, these particles were mixed in deionized water and magnetically stirred. For every suspension, magnetic stirring was performed for 10 h to ensure homogeneous mixing, and then the suspension was dried in an oven at 120 • C for 2 h. After grinding, the dried powders were homogenized with deionized water as a binder, and pressed using a hydraulic press at 3 MPa to form pellets with a diameter of 10 mm and a thickness of 1 mm. The pellets were sintered at 850-1050 • C for 2 h in air. For gas-sensing measurement, a pair of rectangular gold electrodes was formed on a major surface of a sample, through direct-current (DC) magnetron sputtering.

NO 2 -Sensing Measurement
A commercial gas-sensing measurement system (GRMS-215, Partulab Com., Wuhan, China), which has been described in detail in previous papers [31,45], was used for the tests, mainly composed of a 350 mL quartz-sealed chamber and a computer for recording data. The sealed chamber contained four inlet tubes and one exhaust tube. Four inlet tubes were connected to N 2 (99.999%, Zhongxinruiyuan Gas, Wuhan, China), O 2 (99.999%, Zhongxinruiyuan Gas, Wuhan, China), 15 ppm NO 2 -N 2 (97.0%, Foshan Kodi Gas, Foshan, China), and the air, to realize designated internal atmospheres. A flow controller was used to control the gas inflow in every pipe and ensure a stable gas flow. During the response stage, NO 2 , N 2 , and O 2 entered the chamber at specific rates through three tubes, and the total rate was maintained at 300 mL/min. During the recovery stage, ambient air was pumped into the chamber at a rate of 10 L/min. During most measurements, the room temperature was kept at 25 • C and the RH in the air remained around 50%.

Material Characterization
Crystal structures of the prepared samples were investigated by powder X-ray diffraction (BRUKER AXS D8 ADVANCE, Bruker Daltonics, Bremen, Germany), using Cu K α radiation. Morphologies and microstructures of the nanoceramics were characterized by scanning electron microscopy (SEM; SIRION, FEI, Eindhoven, the Netherlands). The distribution of elements was studied by energy-dispersive spectroscopy (EDS, ZEISS Corporation, Jena, Germany), using the OXFORD Aztec 250 instrument (Oxford Instruments, Oxford, UK).

Conclusions
Composite nanoceramics have been prepared through the pressing and sintering of Pt, Au, and SnO 2 nanoparticles. For samples of 1 wt% Pt-SnO 2 , the resistance was greatly increased after being exposed to NO 2 in synthetic air at room temperature. For samples of 5 wt% Au-SnO 2 , the resistance was unusually small in the air, and was dramatically increased by NO 2 in synthetic air at room temperature. Samples of 1 wt% Pt-5 wt% Au-SnO 2 had the strongest response to NO 2 at room temperature, with a response of 875 to 0.5 ppm NO 2 -20% O 2 -N 2 , a response time of 2566 s, and a recovery time of 450 s in the air of 50% RH. Such a room-temperature NO 2 -sensing capability is highly remarkable among those reported in the literature. According to the mechanism study results, it is proposed that NO 2 molecules are chemisorbed on SnO 2 under the catalytic effect of Pt-Au at room temperature, and water molecules in the air have a tendency to desorb NO 2 molecules from SnO 2 through attraction. More studies on composite ceramics of metal oxides and noble metals for sensing oxidizing gases at room temperature are highly desirable.

Conflicts of Interest:
The authors declare no conflict of interest.
Sample Availability: Not available.