Highly sensitive H2S gas sensors based on Pd-doped CuO ...

Accepted Manuscript Title: Highly sensitive H2 S gas sensors based on Pd-doped CuO nanoflowers with low operating temperature Authors: Xiaobing Hu, Zhigang Zhu, Cheng Chen, Tianyang Wen, Xueling Zhao, Lili Xie PII: DOI: Reference:

S0925-4005(17)31202-9 http://dx.doi.org/doi:10.1016/j.snb.2017.06.183 SNB 22654

To appear in:

Sensors and Actuators B

Received date: Revised date: Accepted date:

20-1-2017 5-5-2017 29-6-2017

Please cite this article as: Xiaobing Hu, Zhigang Zhu, Cheng Chen, Tianyang Wen, Xueling Zhao, Lili Xie, Highly sensitive H2S gas sensors based on Pddoped CuO nanoflowers with low operating temperature, Sensors and Actuators B: Chemicalhttp://dx.doi.org/10.1016/j.snb.2017.06.183 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Highly sensitive H2S gas sensors based on Pd-doped CuO nanoflowers with low operating temperature Xiaobing Hu, Zhigang Zhu *, Cheng Chen, Tianyang Wen, Xueling Zhao, Lili Xie School of Environmental and Materials Engineering, College of Engineering, Shanghai Polytechnic University, Shanghai, 201209

*Corresponding author: Fax: +86-21-50215021*8325, Tel: +86-21-50215021*8325; E-mail address: [email protected] (Z. Zhu)

Highlights 

The response (Rg/Ra) of the Pd-doped CuO sensors to 50 ppm H2S was 123.4 at 80 oC, which is much higher than previous reports (Rg/Ra = 3-50). Moreover, when the H2S concentration is as low as 100 ppb, the response of our sensor is still 4.5.



Long recovery time is a common issue for the gas sensors operated at low temperature; we found a solution to apply a short electric current pulse to improve the recovery speed at low temperature.



The Pd doping plays an important role to enhance their response and selectivity toward H2S.

2

Abstract: A facile method was used to prepare Pd-doped CuO nanoflowers with various doping concentrations. The samples were characterized through X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), energy dispersive X-ray spectroscopy

(EDX), inductively

coupled plasma atomic emission spectrometer (ICP-AES), and Brunauer - Emmett Teller (BET) specific surface area analysis. The responses (Rg/Ra or Ra/Rg, where Rg is the resistance in gas, and Ra is the resistance in air) of such sensors exposed to 50 ppm CH4, NO2, C2H5OH, H2S, NH3, and H2 were measured for comparison. For 1.25 wt% Pd-doped CuO nanoflowers, the response (Rg/Ra) to 50 ppm H2S was 123.4 at 80 o

C, which was significantly higher than that of pure CuO (Rg/Ra=15.7). Furthermore,

excellent stability and repeatability of the gas sensor were also demonstrated. The observed results clearly revealed that it is an important and facile approach to detect the H2S at low operating temperature for practical applications. Keywords: CuO; gas sensor; Pd doping; p-type; low temperature

1. Introduction Hydrogen sulfide (H2S) is a colorless, poisonous, flammable, foul odor of rotten eggs, which is usually produced in decomposition of the organic matter, human and animal’s waste, food processing, cooking stove, craft paper mills, tanneries and oil refineries and other industrial activities [1,2]. Exposure to low concentration of H2S may lead to sore throat, cough, eye irritation, etc. H2S is also related to some diseases such as Alzheimer’s disease, Down syndrome, Liver cirrhosis and so on [3,4]. Thus, it is great importance to detect low concentration of H2S effectively for human health. As a typical reducing gas, H2S could be detected by pure or doped metal oxides sensors at high operating temperature [5]. However, it is dangerous to monitor H2S at high temperature for its flammability. Therefore, it is still a great challenge to explore new type of sensors for detecting H2S with low operating temperature. Metal oxide semiconductors (MOS) have been widely used to monitor oxidizing and reducing gases in a simple and cost-effective way [6-8]. The most representative sensing materials are SnO2 [9,10] and ZnO [11], which exhibit n-type semi-conductivity and excellent gas sensitivity, other n-type MOS such as TiO2 [12], WO3 [13], In2O3 [14] and Fe2O3 [15] are being widely selected to monitor various gases as well. So far, a great number of approaches have been conducted to improve 3

gas response of n-type MOS, such as utilizing full electron depletion of sensing materials through nanoparticles [16,17], promoting gas sensing reaction with noble metal or oxide catalysts (chemical sensitization) [18,19] and controlling donor density by various dopants (electronic sensitization) [20]. In contrast for p-type MOS gas sensors such as those CuO, NiO and Co3O4, to date have received relatively little attention, and further systematic investigations are needed for the design of highly sensitive, selective, and reliable p-type gas sensors [21]. CuO is a p-type semiconductor with 1.2 eV band gap [22,23], it is well-known for both scientific research and industries as it has already been widely utilized in lithium ion batteries, super-capacitors, photodetectors, solar cells, heterogeneous catalyst, magnetic storage media and field emissions [24,25]. Some groups have paid attention to the gas sensing characteristics of such doped or undoped CuO. Ramgir et al. [26] first demonstrated room temperature sub–ppm (100–400 ppb) H2S sensing using CuO thin films, and found that in the 100-400 ppb concentration range, the response of CuO films for H2S was highly reversible, however, at higher concentrations, the response curves turn irreversible. Chao et al. [27] reported that the response of the CuO nanorads sensor to ethanol (1000 ppm) was 9.8 when operating at 210 oC. Liu et al. [28] reported the sensor with wormlike CuO structures showed excellent sensor performances toward ethanol and methanol detection. Aslani et al. [29] reported that the CuO nanostructures with cloudlike morphology exhibited higher sensitivity to CO gas, moreover, the sensitivity and selectivity were better than that of the pure CuO. To improve their gas sensing properties including selectivity, sensitivity, response time and operating temperature, metal oxides can be modified either on the surface or lattice with different materials [30-32]. Metal oxide loaded with catalytic metals or noble metals nanoparticles on the surface have commonly been used. For instance, Kang et al. reported that 2.2 wt% Cr added to CuO nanostructures by solvothermal reaction significantly increased the responses to 100 ppm NO2 from 7.5 to 134.2 [33]. Barreca et al. reported that CuO-TiO2 nanocomposites functionalized with Au nanoparticles exhibited excellent gas sensing properties [34]. Kim et al. [35] prepared CuO one-dimensional nanostructures functionalized with Pd, and indicated that the Pd-functionalized CuO nanorod sensors showed a high response of 167.0 at 50 ppm, however, the operating temperature was relatively high (300 oC). Low operating temperature is able to reduce the power consumption, and it is beneficial for the large-scale use of the gas sensors. Qin et al. reported that the W18O49/CuO core-shell 4

nanorods showed satisfactory sensing response to 0.1 - 1 ppm NO2 gas with excellent dynamic response-recovery characteristics at 50 oC [36]. Xu et al. reported that the CuO-nanoparticles/ZnO-nanords heterostructure exhibited higher response to 50 ppm TEA gas, operating as low as 40 oC [37]. Unfortunately, sensing performances, like sensitivity, response time and recovery time of these sensors at low operating temperature still need further enhance to satisfy the criterion for practical applications. Therefore, it remains great need to develop high performance gas sensors in mild conditions. In this work, a high performance Pd-doped CuO sensor was fabricated by a simple water bath heating route, which is able to detect H2S at low operating temperature (the operating temperature of the sensor was varied between 70 oC and 340 oC).The Pd dopants play an important role to enhance their response and selectivity toward H2S, and the mechanism for such tremendous enhancement is thoroughly studied in this paper.

2. Experimental 2.1 Preparation of CuO All chemicals were purchased with reagent grade from Sinopharm Chemical Reagent Co. Ltd, and used without further purification. CuO nanoflowers were prepared by the following procedure: 24.16 g Cu(NO3)•3H2O was dissolved in 1000 mL deionized water to prepare 0.1 mol/L copper nitrate solution. Then, 12.62 g hexa-methylenetetramine was poured into 20 mL above solution with continuous stirring for 1 h to get bright blue solution. The mixed solution was stirred for 90 min and the water bath was set at 80 oC. During the reaction, the bright blue Cu-hydroxide precipitate converted into dark brown CuO. After cooling down to room temperature, the CuO mixture were separated by centrifugation at 8000 rpm for 10 min followed by washing three times with deionized water and ethanol, respectively, and then dried at 80 oC in an oven. 2.2 Pd-Doped CuO Nanoflowers Pd-doped CuO nanoflowers were prepared by the following procedure: 2 mmol/L palladium precursor solution (H2PdCl4,) was prepared by adding 0.0887 g PdCl2 into 6 mL HCl (0.2 mol) and then diluted into 250 ml with deionized water. After that, 20 mL copper nitrate solution and H2PdCl4 was mixed together, the remaining steps are the same as mentioned above. For a convenient comparison, we define the doping of 5

different concentration of Pd as 0.00Pd-CuO, 0.75Pd-CuO, 1.25Pd-CuO and 1.50Pd-CuO, respectively, which means the Pd doping concentration were 0.00 wt%, 0.75 wt%, 1.25 wt% and 1.50 wt%, respectively. 2.3 Characterization The phase and crystallinity of the CuO specimens were analyzed by X-ray diffraction with a monochromatized Cu target radiation resource (λ=1.5418Å)， scanning from 10 - 80 degree (XRD, D8-Advance, Bruker, Germany). The morphology of the powders was investigated by field emission scanning electron microscopy (FE-SEM, SU8220, Hitachi Co. Ltd, Japan) at an accelerating voltage of 10 kV. Energy Dispersive X-Ray spectroscopy (EDX, S-4800, Hitachi Co. Ltd, Japan) was used to confirm the composition of the CuO nanostructures. Transmission electron microscopy (TEM, JEM-2100F, JEOL Co. Ltd, Japan) was used to examine microstructure of the CuO nanostructures under a working voltage of 200 kV. The surface areas were measured by the Brunauer-Emmett-Teller method (BET, Tristar II 3020, Micromeritics Instrument Co Ltd. USA). The concentrations of Pd-doped specimens were determined by inductively coupled plasma atomic emission spectrometer (ICP-aes, A-6300, Thermo Scientific, USA). 2.4 Gas Sensing Characteristics The products were mixed with deionized water at a weight ratio of 4:1 to form a paste. The sensors were made by coating ceramic tube with the paste to form a 10 μm sensing film and dried under infrared radiant heater. A pair of gold electrodes was previous printed at each end of the ceramic tube before it was coated with the paste, the schematic structure of the ceramic electrode is shown in Fig. 1. The sensing element was heat treated at 500 oC for 2 h to remove the bound water and generate dense oxide film on the surface of the gas sensor. After cooling down, a Ni-Cr heating wire was inserted into the tube to form an indirect-heated gas sensor. The sensor was then placed in an aging instrument and the temperature of the aging was stabilized at 400 oC for 7 days. The electrical properties of the sensor were measured using a Gas Sensing Analysis System (WS-30A , Zhengzhou Weisheng Tech Co, Ltd). During the testing, a given amount of target gases mixed with dry air were injected into a 15 L chamber to obtain desired concentrations. The sensor response was defined as the ratio (S = Rg/Ra) of the resistance of the sensor in air (Ra) to that in target gases (Rg). The operating 6

temperature of the sensor was varied between 70 oC and 340 oC. The response and recovery time were defined as the time taken by the sensor to achieve 90% of the total resistance change in the case of adsorption and desorption, respectively. Preferred place for Figure 1

3. Results and discussion 3.1 Structure and morphology of sensing materials Preferred place for Figure 2

Fig. 2 illustrates XRD patterns of 0.00 - 1.50 wt% Pd-doped CuO nanoflowers. The patterns of CuO are specified as monoclinic CuO (JCPDS No. 72-0629), and shows a well-defined diffraction peaks at 2θ of 32.61°, 35.59°, 38.70°, 48.84°, 53.65°, 58.22°, 61.44°, 66.32° and 68.03°, corresponding to the (110), (-111), (111), (-202), (020), (202), (-113), (-311) and (220) planes of monoclinic CuO. All XRD patterns have a similar view at first sight as seen in Fig. 2 (a-d). However, a closer look at the specific regions (35 - 40°) of XRD patterns reveals the differences. Comparing the Pd-doped CuO patterns with the pure CuO, the (-111) and (111) peaks were slightly moved towards the lower angle with the increase of Pd dopant from 0.75 wt% to 1.50 wt%, which implies the change of lattice constant after Pd doping. For this phenomenon, the ionic radius of Pd2+(0.86 Å) is larger than that of Cu2+(0.73 Å), the substitution of Cu2+ by Pd2+ ions induces lattice expansion and thus causes the peak moving to lower angle. This clearly identifies that the Pd2+ ions were successfully doped into the CuO matrix. Preferred place for Figure 3

EDX analysis was employed to further confirm the composition of the CuO nanoflower. Figure 3 shows the EDX spectra of 1.25 wt% Pd-doped CuO nanoflower. The peaks associated with O, Cu and Pd are clearly presented in Fig. 3. The other peaks, such as Au, were ascribed to the conductive coating, and it thus proves Pd2+ ions doped into the CuO. Meanwhile, the concentrations of the 0.75Pd-CuO, 1.25Pd-CuO and 1.50Pd-CuO specimens were also determined by inductively coupled plasma mass spectroscopy, the concentration for Pd were 0.27 at%, 0.76 at% and 0.83 at%, respectively. This further confirms the Pd2+ ions were successfully doped into the CuO matrix. 7

Preferred place for Figure 4

The FESEM morphologies of Pd-doped CuO nanostructures doped with different concentration of Pd are shown in Fig. 4. As exhibited in Fig. 4 (a), the pure CuO nanostructures have a spindle-like morphology and the particle diameter ranged from 200 nm to 4 μm, showing poor uniformity. However, as shown in Fig. 4 (b-c), the morphology changes from spindle-like into flower-like nanostructures after increasing Pd concentration. When Pd doping concentration is 1.25 wt% (Fig. 4 (c)), 3D flower-like CuO with hierarchical nanostructures is thus obtained, the diameter of the CuO nanoflowers is ca. 400 nm and they are homogeneous distributed. It is obviously that the flower-like CuO is consisted of nanosheets with the size around 200 nm. Pd doping plays an important role during the synthesis of CuO, and the resulting growth-mechanism is referred from the preparation of relative metal oxides [38]. During the process, copper nitrate solution produces Cu2+ ions to form a blue precipitate of copper hydroxide Cu(OH)2. After increasing temperature, a blue color of the parent precipitate gradually turned into dark brown indicating dehydration of Cu(OH)2 and finally forming the nano-CuO. Once Pd was introduced, it did not affect the precipitate color, but seemed to have a consecutive influence on the growth rate of the nanoflower CuO. The Pd ions tend to be adsorbed on the CuO basal planes and thus modify its chemical surface properties, such as relative surface free energy on the facets, and consequently, it results into blocking and preventing the incorporation of other molecules from solution into the CuO crystal lattice and slow down the growth in such particular direction. The presence of higher amounts of Pd also influences dehydration of (CuOH)2 and retards Cu(OH)2 to CuO phase transformation. Further increasing the Pd to 1.50 wt%, the morphology of the CuO does not significantly altered, but the uniformity of the product declines again. The morphologies of the pure CuO and 1.25Pd-CuO nanostructures were then further characterized by TEM. Fig. 4 (e) shows the TEM images of the pure CuO, it confirms the result observed in FESEM. Fig. 4(f) clearly indicates that 1.25Pd-CuO nanoparticles have a flower-like morphology. It is noteworthy that there are plenty of fluffs between nanosheets and surrounding flower-like nanoparticles, those fluffs are able to provide enormous defects and chemical active sites for the gas molecule adsorption. Preferred place for Figure 5 8


The BET surface areas of the pure CuO and 1.25Pd-CuO were investigated by nitrogen adsorption measurement (as shown in Fig. 5). The BET surface area of the 1.25Pd-CuO is 25.3 m2/g, which is 1.8 folds larger than pure CuO (14.0 m2/g). Moreover, no surface porosity is found in both samples. The larger surface area of the 1.25Pd-CuO can be beneficial for the enhancement of gas sensing properties. 3.2 Sensing properties The operating temperature plays an important role for gas sensors, since it has great influence on the adsorption and desorption during the gas-sensing process. Therefore, we first implemented experiments operating between 70 oC and 340 oC to obtain optimized operating temperatures of the four sensors. Then, we analyzed and identified the optimal composition by comparing series of samples. As can be seen from the Fig. 6(a), the responses of the four sensors to 50 ppm H2S all exhibit a maximum value at 80 oC and decreases to 1.1 - 1.6 as the operating temperature increases to 340 oC. The doping of 1.25 wt% Pd reaches the maximum H2S responses at 80 oC, and the value is around 123.4 at 80 oC, which is dramatically higher than other three sensors. For example, the H2S response of the 1.25Pd-CuO sensor is 7.9 times higher than that of pure CuO sensor (Rg/Ra=15.7) at 80 oC. As the doping ratio increased to 1.50 wt%, the response of the sensor decreases to 61.3, but still higher than pure CuO sensor. When the operating temperature further increased, the response differences of all sensors reduces to less than 10 once operating temperatures are higher than 200 oC, as shown in the inset of Fig. 6(a). The Pd-doped sensors exhibit extremely low response (1.1-1.5) as well as the pure CuO sensor at high operating temperature (above 300 oC), this is mainly own to the high intrinsic carrier concentration of the CuO matrix at high temperatures [39]. Therefore, the best doping ration is 1.25 wt% Pd and operating at 80 oC. The beauty of low operating temperature has advantages to prevent grain growth and greatly reduce energy consumption, which is urgently expected for real application. Moreover, we have also tested the water-sensing response of the sensor exposing to different relative humidity water vapor as show in Fig, 6(b). The atmosphere of RH was prepared with different saturated salt solutions in their equilibrium states including LiCl for 11%RH, MgCl2 for 33%RH, Mg(NO3)2 for 54%RH, NaCl for 75%RH, KCl for 85%RH, and KNO3 9

for 95%RH [40]. As can be seen from Fig. 6(b), the highest response of the 1.25Pd-CuO sensor to 95%RH water vapor is about 1.1, while the responses are all less than 1.2. Thus, the humidity effect of the sensing characteristics of the sensor is negligible. Selectivity is also a key parameter to gas sensors. The sensing performances were examined to 50 ppm CH4, NO2, C2H5OH, H2S, NH3, and H2 at 80 oC, as shown in Fig. 7. For the pure CuO, the response to 50 ppm H2S and NH3 are 15.7 and 1.9, respectively, while the responses to other gases ranging between 1.1 and 1.5. After doping 0.75 wt% Pd, the response to 50 ppm H2S increases to 95.2 while the responses to the other gases remain similar. When doped with 1.25 wt% Pd, the response to 50 ppm H2S reaches its highest value 123.4. Although the response to 50 ppm NH3 is also increased, the enhancement of the NH3 response is only 1.5 folds, which is negligible compared to that of H2S response (123.4/15.7=7.9). Further increased the doping content to 1.50 wt%, the response to H2S reduces into 60.2. Comparing to the other sensors, for example, Wang at al. [41] reported that the response of the CuO-NiO core-shell microspheres sensors to 100 ppm H2S was just 47.6, so the 1.25 wt% Pd doping to the CuO nanostructures resulted in highest selective and sensitive detection of H2S at low operating temperature, i.e. 80 oC. Preferred place for Figure 7

Response and recovery times are important factors for gas sensors, fast response and recovery is able to lead a real-time detection hazard gas. Fig. 8 shows the dynamic response-recovery curves of the sensor based on 1.25Pd-CuO to 50ppm H2S at 80 oC. After the sensor exposed to H2S gas, it takes less than 15 s to reach 90% of the resistance variation. Although the sensors based on Pd-doped CuO are highly sensitive and fast response to detect H2S gas, an extremely long recovery time was noticed during the process. Fig. 8 also shows that it takes more than 3500 s to recover about 8% of the maximum value after removing the H2S gas. In fact, the response could not be completely recovered to the baseline while operating at 80 oC. Actually, it is a common issue for the gas sensor operated at low temperature. When the Pd-doped CuO sensor is exposed to H2S gas, a CuS layer might form on the surface of the CuO nanoflowers according to the following spontaneous chemical reaction: CuO(s) + H2S(g) = CuS(s) + H2O(g) 10

(1)

It was reported that, the change in the Gibbs free energy for the sulfuration reaction of Eqn. (1) is -119.46 kJ/mol (at 20 oC) and -114.1 kJ/mol (at 80 oC) [35]. The calculated thermodynamic data indicates that the reaction could spontaneously occur and is favored at low temperature. This explains the ability of CuO nanoflowers to respond to H2S at low temperature. However, copper (II) sulfide (CuS) is unstable, and can be slowly oxidized in the air as shown in Eqn. (2). What’s more, CuS is prone to be transformed into CuO and SO2 at higher temperatures (＞220 oC) as shown in Eqn. (3). As a result, the gas sensor could thus make a fast recovery after heating at 300 oC. CuS(s) + O2(g) → 2CuO(s) + 2S(s) 2CuS(s) + 3O2(g)

2CuO(s) + 2SO2(g)

(2) (3)



In order to improve the recovery speed at low temperature while maintaining high response, we found a solution to applying a short electric current pulse to accelerate the detachment of the H2S molecules from the CuO surfaces using the Joule heating effect, as shown in the insert of Fig. 9(a). Here, Rl is a load resistor, and the voltage drop on Rl is denoted as Vout, Vc is the test voltage and Vh is the heating voltage applied on the Ni-Cr alloy resistor, Vm is the 4.6V electric modulation pulse voltage. We found that 1.25Pd-CuO sensors can recover quickly after heating at 300 oC, and the heating time should be optimized since incomplete desorption of the sensors occurred if the time is too short. Meanwhile, the lifetime of the sensors will be shortened if the heating treatment is too long, which also waste much energy. We tested the response of the sensor to 1 ppm H2S gas. A heating pulse (300 oC) is selected to make a fast recovery when the response reached the first balance, while the response of the 1.25Pd-CuO sensor to 1 ppm H2S gas is 63.8. Fig. 9(a) shows the response of the sensor to 1 ppm H2S gas with different heating time during recovery process. The response is 34.6, 58.1 and 63.8 when the heating time is 10 s, 50 s and 80 s, respectively. It revealed that the sensor remains normal function when the heating treatment is more than 80 s. Fig. 9(b) shows the dynamic sensing transients of 11

the 1.25Pd-CuO sensor to 1 ppm H2S at 80 oC, the response time is 78 s even the sensor was exposed to 1 ppm H2S, and the response (Rg/Ra) is ranged from 63.8 to 69.0. Once H2S was removed, the response could not be recovered to its baseline. Then, we applied a 4.6 V (the temperature was 300 oC) voltage pulse, and the 1.25Pd-CuO sensor demonstrates a fast recovery within 12 s. Reproducibility of the sensor is the major concern for the practical use of the sensor, Fig. 9(b) also shows the reproducibility of the four randomly selected sensors (sensor (a-d)) on successive exposure to and deviate from (5 cycles) 1 ppm of H2S gas, the result indicates that the sensors based on Pd-doped CuO exhibits good tolerance, no obvious resistance attenuation is detected, and the sensors maintain its gas-sensing performance such as response/recovery speed during the process. Thus, it clearly shows that the sensors based on Pd-doped CuO have a very stable sensing characteristic. In addition, such a “recovery heating” process had tiny effect on the response of the sensor, which makes the gas sensor highly recyclable. Therefore, with a recovery modulation voltage, we have developed high performance H2S gas sensors based on the Pd-doped CuO at low operating temperature. Fig. 10(a) shows the relationship between the response of 1.25Pd-CuO sensor and Preferred Figure 10 o different H2S concentrations at 80 place C. Itfor can be observed that gas response increased

approximately linearly with H2S concentration in the interval of concentrations from 100 ppb to 1 ppm. In addition, the response slowly changes after the H2S concentration is higher than 10 ppm, which indicates that the sensors become more or less saturated. Long-term stability of 1.25Pd-CuO sensor is shown in Fig. 10(b), sensing performances were measured for 6 cycles, and every cycle was 7 days. The results reveal that the response of sensors to 50 ppm H2S have no obvious degradation after being kept in the air for 42 days, which indicates that the sensor structure was robust and it is a promising sensor for detecting the H2S at low operating temperature with low power consumption.

3.3 The sensing mechanism Pure CuO is a typical p-type semiconductor oxide, and its gas-sensing mechanism can be explained by the change of resistance caused by the adsorption/desorption and reaction of gas molecules on the surface of the semiconductor [42]. When p-type CuO 12

semiconductor is exposed to air, oxygen molecules are adsorbed on the surface of the sensor and ready to be ionized by electrons to form adsorbed oxygen ions (O2-, O-, O2-), which were shown in Eqn (4) - (7). In this process, the capture of electrons leads to the accumulation of holes near the CuO surface, as a result, decreasing the resistance of the sensor. Once the sensor exposed to H2S (reducing gas) atmosphere at a moderate temperature, the H2S molecules are able to react with the adsorbed oxygen ions and thus release electrons [43], as shown in Eqn (8) - (9). Consequently, the released electrons will recombine with holes and resulting in a decreased hole concentration. The resistance of the CuO-based sensors is thus increased upon contacting with the H2S gas. Once H2S is removed by introducing air, the sensor is “washed” and the number of free electrons is reduced. The above mechanism leads to a reversible H2S sensing process for CuO sensors [44]. O2(gas) → O2(ads)

(4)

O2(ads) + e- → O2-(ads)

(5)

O2-(ads) + e- → 2O-(ads)

(6)

O-(ads) + e-→ O2-(ads)

(7)

H2S(g) + 3O2-(ads) → H2O(g) + SO2(g) + 6e-

(8)

H2S(g) + 3O-(ads) → H2O(g) + SO2(g) + 3e-

(9)


As shown in Fig. 11, the sensing performance could be enhanced by doping process. Two types of mechanisms, the electronic and the chemical sensitization, have been previous reported [45-46]. Here, the improved response for Pd-doped CuO is mainly attributed by electronic sensitization [47]. As we know, PdO is also a p-type semiconductor with high intrinsic carrier concentration and conductivity. Compared to other materials, it has lower resistance and better conductivity even at room temperature (25 oC). This is one of the main reasons why Pd doping could reduce the operating temperature for Pd-doped CuO sensor. In addition, the catalytic ability of PdO is able to promote the sensing material to capture oxygen molecules and form absorbed oxygen [48]. Pd doping help CuO to capture H2S molecules around the surface. We also studied the effect of Pd doping on the electrical properties of the CuO nanoflowers with various doping concentration (0, 0.75, 1.25 and 1.50 wt%). After PdO was incorporated on the surface of CuO, the sensor resistance will not changed. Indeed, the average Ra value determined from 4 different sensors with 13

various doping concentration were 0.2 ± 0.05 KΩ. The resistance of the same doping sensors did not show notable differences when they exposed to air. This indicates that the significant enhancement of the H2S response does not emanate from the Ra value variation by Pd-doping. Meanwhile, according to the electronic sensitization, PdO molecules on CuO surface can participate in the reaction and released electrons when the sensor was exposed to the H2S gas. As a result, the released electrons will recombine with holes and leading to a decreasing of hole concentration, the resistance of the Pd-dopant CuO sensor is thus increased upon contacting with the H 2S gas. And the response of the sensor was defined as Rg/Ra, so the response increases with the increase of Rg. On the other hand, compared to pure CuO, 1.25Pd-doped CuO nanoflowers have larger BET surface and a higher density of adsorption sites, which results in the considerable response even at relatively low gas concentrations. However, as the doping concentration increased to 1.50 wt%, the surface disorder occurs and it is inevitably accompanied by the increase of surface states density, which might lead to the pining of surface Fermi level and the decrease of sensor response [49]. In a word, the doping of Pd is mainly attributed to the enhanced gas response of CuO based sensors. A further detailed sensing mechanism of Pd-doped CuO nanostructures is still under investigation in our group. A comparison between the H2S-Sening performances of the various metal-oxide semiconductor sensor and literature is summarized in Table 1 [26,35,50-53]. The performance of the Pd-doped CuO sensor is comparable with or higher than that of the recently developed sensors.

Table. 1 Gas responses to H2S of the sensors in the present study and those reported in the literature

14

a

Calculated data from Response=(Rg-Ra)/Ra*100% The lowest test concentration in paper, but the accurate limit of detection not given.

b

4. Conclusion Sensors based on Pd doped CuO nanoflowers have been prepared by a simple water bath heating method for H2S gas detection. All characterization results show that the Pd2+ ions have successfully been doped into the CuO matrix. The doping of Pd plays substantial role for enhanced sensing performance to H2S gas. After doping 1.25 wt% Pd, the sensors reaches the maximum H2S responses (Rg/Ra=123.4) at 80 oC, and it takes less than 15 s to reach 90% of the maximum value. After applying a 4.6 V voltage pulse, the recovery time can be shortened within 12 s. Moreover, when the H2S concentration is as low as 100 ppb, the response of our sensor is still 4.5. Such 1.25Pd-CuO sensors have excellent reversibility, long-term stability and selectivity to H2S gas at low operating temperature, and it is an ideal candidate for H2S detection.

Acknowledgements This work was supported by the National Natural Science Foundation of China (No. 61471233, 21504051), the Program for Professor of Special Appointment (Eastern Scholar) at SIHL, Shuguang Project and Young Teacher Training Project of Shanghai University supported by Shanghai Municipal Education Commission and Shanghai Education Development Foundation (Nos. 14SG52 and ZEGD15012), the Key Subject of Shanghai Polytechnic University (MST, XXKZD1601).

References 15

[1] Y. Y. P. Mok, M. Atan, C. Y. Ping, W. Z. Jing, M. Bhatia, S. Moochhala, P. K. Moore, Role of hydrogen sulphide in haemorrhagic shock in the rat: protective effect of inhibitors of hydrogen sulphide biosynthesis, Br. J. Pharmacol. 143 (2004) 881-889. [2] B. Teague, I. Asiedu, P. K. Moore, The smooth muscle relaxant effect of hydrogen sulphide in vitro: evidence for a physiological role to controlintestinal contractility, Br. J. Pharmacol. 137 (2002) 139-145. [3] M.N. Bates, J. Crane, J. R. Balmes, N. Garrett, Investigation of hydrogen sulfide exposure and lung function, asthma and chronic obstructive pulmonary disease in a geothermal area of New Zealand, PLoS One 10 (2015) 122062-122078. [4] P.A. Hessel, F.A. Herbert, L.S. Melenka, K. Yoshida, M. Nakaza, Lung health in relation to hydrogen sulfide exposure in oil and gas workers in Alberta, Canada, Am. J. Ind. Med. 31 (1997) 554-557. [5] Y.Y. Wang, G.T. Duan, Y.D. Zhu, H.W. Zhang, Z.K. XU, Z.F. Dai, W.P. Cai, Room temperature H2S gas sensing properties of In2O3 micro/nanostructured porous thin film and hydrolyzation-induced enhanced sensing mechanism. Sens. Actuators B 228 (2016) 74-84 [6] N. Yamazoe, Toward innovations of gas sensor technology. Sens. Actuators B 108 (2005) 2-14 [7] M.E. Franke, T.J. Koplin, U. Simon, Metal and metal oxide nanoparticles in chemiresistors: Does the nanoscale matter? Small. 2 ( 2006) 36-50. [8] Y. Shimizu, M. Egashira, Basic aspects and challenges of semiconductor gas sensors. MRS Bull. 24 (1999) 18-24. [9] E. Comini, G. Faglia, G. Sberveglieri, Z. Pan, Z.L. Wang, Stable and highly sensitive gas sensors based on semiconducting oxide nanobelts, Appl. Phys. Lett. 81 (2002) 1869-1871. [10] A. Kolmakov, Y. Zhang, G. Cheng, M. Moskovits, Detection of CO and O2 using tin oxide nanowire sensors, Adv. Mater. 15 (2003) 997-1000. [11] Z. Jing, J. Zhan, Fabrication and gas-sensing properties of porous ZnO nanoplates, Adv. Mater. 20 (2008) 4547-4551. [12] I.D. Kim, A. Rothschild, B.H. Lee, D.Y. Kim, S.M. Jo, H.L. Tuller, Ultrasensitive chemiresistors based on electrospun TiO2 nanofibers, Nano Lett. 6 (2006) 2009-2013. [13] H.G. Moon, Y.S. Shim, D.H. Kim, H.Y. Jeong, M. Jeong, J.Y. Jung, S.M. Han, J.K. Kim, J.S. Kim, H.H. Park, J.H. Lee, H.L. Tuller, S.J. Yoon, H.W. Jang, Self-activated ultrahigh chemosensitivity of oxide thin film nanostructures for transparent sensors, Sci. Rep. 2 (2012)588. [14] D. Zhang, Z. Liu, C. Li, T. Tang, X. Liu, S. Han, B. Lei, C. Zhou, Detection of NO2 down to ppb levels using individual and multiple In2O3 nanowire devices, Nano Lett. 4 (2004) 1919-1924. [15] H.J. Kim, K.I. Choi, A. Pan, I.D. Kim, H.R. Kim, K.M. Kim, C.W. Na, G. Cao, J.H. Lee, Template-free solvothermal synthesis of hollow hematite spheres and their applications in gas sensors and Li-ion batteries, J. Mater. Chem. 21 (2011) 6549-6555. 16

[16] C.N. Xu, J. Tamaki, N. Miura, N. Yamazoe , Grain size effects on gas sensitivity of porous SnO2-based elements. Sens Actuators B 3 (1991) 147-155 [17] T. Kida, T. Doi, K. Shimanoe , Synthesis of monodispersed SnO2 nanocrystals and their remarkably high sensitivity to volatile organic compounds. Mater. 8 (2010) 2662-2667 [18] D.J. Yang, I. Kamienchick, D.Y. Youn, A .Rothschild, I.D. Kim ,Ultrasensitive and highly selective gas sensors based on electrospun SnO2 nanofibers modified by Pd loading. Adv. Funct. Mater. 20 (2010) 4258-4264 [19] N. Yamazoe, Toward innovations of gas sensor technology, Sens Actuators B 108 (2005) 2-14 [20] N. Yamazoe, K. Shimanoe, New perspectives of gas sensor technology. Sens Actuators B 138 (2009) 100-107 [21] H.J. Kim, J.H. Lee, Highly sensitive and selective gas sensors using p-type oxide semiconductors:overview. Sens. Actuators B 192 (2014) 607-627 [22] A. Tricoli, M. Righettoni, A. Teleki, Semiconductor gas sensor: dry synthesis and application. Angew. Chem. Int. Ed. 49(2010) 7632-7659 [23] J. Demel, A. Zhigunov, I. Jirka, M. Klementová, K. Lang, J. Coll. Facile synthesis of CuO nanosheets via the controlled delamination of layered copper hydroxide acetate. Interf. Sci. 452 (2015) 174-179 [24] T. Jiang, Y. Wang, D. Meng, M. Yu, Facile synthesis and photocatalytic performance of self-assembly CuO microspheres. Superlatt. Microst. 85 (2015) 1-6 [25] Y. Wang, T. Jiang, D. Meng, H. Jin, M. Yu, Controllable fabrication of nanowire-like CuO film by anodization and its properties. Appl. Surf. Sci. 349 (2015) 636-643 [26] N.S. Ramgir, S. Kailasa Ganapathi, M. Kaur, N. Datta, K.P. Muthe, D.K. Aswal, S.K. Gupta, J.V. Yakhmi, Sub-ppm H2S sensing at room temperature using CuO thin films. Sens Actuators B 151 (2010) 90-96 [27] C. Yang, X. Su, F. Xiao, J. Jian, J. Wang, Gas sensing properties of CuO nanorods synthesized by a microwave-assisted hydrothermal method. Sens. Actuators B 158 (2011) 299-303 [28] X. Liu, J. Zhang, Y. Kang, S. Wu, S. Wang, Brochantite tabular microspindles and their conversion to wormlike CuO structures for gas sensing. CrystEngComm 14 (2012) 620-625 [29] A. Aslani, V. Oroojpour, CO gas sensing of CuO nanostructures, synthesized by an assisted solvothermal wet chemical route. Physica B 406 (2011) 144-149. [30] E. Comini, C. Baratto, I. Concina, G. Faglia, M. Falasconi, M. Ferroni, Metal oxide nanoscience and nanotechnology for chemical sensors, Sens Actuators B 179 (2013) 3-20. [31] I. Hayakawa, Y. Iwamoto, K. Kikuta, S. Hirano, Gas sensing properties of platinum dispersed-TiO2 thin film derived from precursor, Sens Actuators B 62 (2000) 55-60. [32] K. Zakrzewska, M. Radecka, M. Rekas, Effect of Nb, Cr, Sn additions on gas sensing properties of TiO2 thin films, Thin Solid Films. 310 (1997) 161-6. 17

[33] K.M. Kim, H.M. Jeong, H.R. Kim, K.I. Choi, H.J. Kim, J.H. Lee, Selective detection of NO2 using Cr-doped CuO nanorods, Sensors 12 (2012) 8013-8025. [34] D. Baghriche, S. Rtimi, G. Pulgarin, R. Sanjines, J. Kiwi, Novel Synthesis and gas sensing performances of CuO–TiO2 nanocomposites functionalized with Au nanoparticles, J. Phys. Chem. C 115 (2011) 10510-10517 [35] H. Kim, C. Jin, S. Park, S. Kim, C.M. Lee, H2S gas sensing properties of bare and Pd-functionalized CuO nanorods, Sens Actuators B 161 (2012) 594-599 [36] Y.X. Qin, X.J. Zhang, Y. Liu, W.W. Xie, Highly aligned array of W 18O49/CuO core-shell nanorods and its promising NO2 sensing properties, J. Alloys Compd. 673 (2016) 364-371 [37] Q. Xu, D.X. Ju, Z.C. Zhang, S.Yuan, J. Zhang, H.Y. Xu, B.Q. Cao, Near room-temperature trimethylamine sensor constructed with CuO/ZnO P-N heterostructural nanorods directly on flat electrode . Sens. Actuators B 225 (2016) 16-23 [38] V. Petr, Š. Václav, H. Jirˇí, K. Martin, Shape-controlled synthesis of Sn-doped CuO nanoparticles for catalytic degradation of Rhodamine B, J. Colloid Interface Sci. 481 (2016) 28–38 [39] W.W. Yu, Y. San, T.N. Zhang, K.N. Zhang, S.X. Wang, X. Chen, N. Dai, CuO/WO3 hybrid nanocubes for high-responsivity and fast-recovery H2S sensors operated at low temperature, Part. Part. Syst. Charact 33 (2016) 15-20 [40] T. Fei, H.R. Zhao, K. Jiang, T. Zhang, Synthesis and humidity sensitive property of cross-linked water-resistant polymer electrolytes, Sens. Actuators B 208 (2015) 277-282 [41] Y.F. Wang, F.D. Qu, J. Liu, Y. Wang, J.R. Zhou, Enhanced H2S sensing characteristics of CuO-NiO core-shell microspheres sensors, Sens. Actuators B 209 (2015) 515-523. [42] C. Yang, F. Xiao, J. Wang, X.T. Su, 3D-flower and 2D sheet-like CuO nanostructures: microwave-assisted synthesis and application in gas sensors, Sens. Actuators B 207 (2015) 177-185. [43] X.H. Liu, J.Zhang, Y.F. Kang, S.H. Wu, S.R. Wang, Brochantite tabular microspindles and their conversion to wormlike CuO structures for gas sensing, CrystEngComm. 14 (2012) 620-625. [44] Y.F. Wang, F.D. Qu, J. Liu, Y. Wang, J.R. Zhou, S.P. Ruan, Enhanced H 2S sensing characteristics of CuO-NiO core-shellmicrospheres sensors, Sens. Actuators B 209 (2015) 515-523. [45] Y.L. Liu, H. Wang, Y. Yang, Z.M. Liu, H.F. Yang, G.L. Shen, R.Q. Yu, Hydrogen sulfide sensing properties of NiFe2O4 nanopowder doped with noble metals, Sens. Actuators B 102 (2004) 148-154. [46] Y.D. Zhang, Z. Zheng, F.L. Yang, Highly sensitive and selective alcohol sensors based on Ag-doped In2O3 coating, Ind. Eng. Chem. Res 49 (2010) 3539-3543 [47] S.H. Wei, Y.Yu, M.H. Zhou, CO gas sensing of Pd-doped ZnO nanofibers synthesized by electrospinning method, Materials Letters 64 (2010) 2284-2286 [48] A. Kolmakov, D.O. Klenov, Y. Lilach, Enhanced gas sensing by individual SnO2 nanowires and nanobelts functionalized with Pd catalyst particles. Nano Letters, 18

5 (2005) 667-673. [49] G. Korotcenkov, I. Boris, V. Brinzari, S.H. Han, B.K. Cho, The role of dopingeffect on the response of SnO2-based thin film gas sensors: analysis based onthe results obtained for Co-doped SnO2 films deposited by spray pyrolysis, Sens. Actuators B 182 (2013) 112-124. [50] S.J. Kim, C.W. Na, I.S. Hwang, J.H. Lee, One-pot hydrothermal synthesis of CuO–ZnO composite hollow spheres for selective H 2 S detection, Sens. Actuators B 168 (2012) 83–89 [51] S. Sonia, P.S. Kumar, N.D. Javram, Y. Masuda, D. Mangalaraj, C.M. Lee, Superhydrophobic and H2S gas sensing properties of CuO nanostructured thin films through a successive ionic layered adsorption reaction process, RSC Adv 6 (2016) 24290-24298 [52] Y. Ni, P. Deng, Y. Zhao, P. Wang, L. Xing, Y. Zhang, X. Xue, The conversion of PN-junction influencing the piezoelectric output of a CuO/ZnO nanoarray nanogenerator and its application as a room-temperature self-powered active H2S sensor, Nanotechnology 25 (2014) 265501. [53] K. Choi, H. Kim, Y. Kang, J.H. Lee, Ultraselective and ultrasensitive detection of H2S in highly humid atmosphere using CuO-loaded SnO2 hollow spheres for real-time diagnosis of halitosis, Sens Actuators B 194 (2014) 371–376 [54] N.Y. Zhao, Z. Chen, W. Zeng, Enhanced H2S sensor based on electrospun mesoporous SnO2 nanotubes, J Mater Sci: Mater Electron 26 (2015) 9152–915

19

Figure Captions Fig. 1.

Schematic structure of the gas sensor.

Fig. 2. X-ray diffraction (XRD) pattern of the Pd-doped CuO specimens: (a) 0.00Pd-CuO, (b) 0.75Pd-CuO, (c) 1.25Pd-CuO, (d) 1.50Pd-CuO. Fig. 3. EDX spectra of 1.25 wt% Pd-doped CuO nanoflower. Fig. 4. FESEM images of : (a) 0.00Pd-CuO, (b) 0.75Pd-CuO, (c) 1.25Pd-CuO, (d) 1.50Pd-CuO; TEM images of (e) 0.00Pd-CuO, (f) 1.25Pd-CuO. Fig. 5. Typical N2 adsorption-desorption isotherms of different specimens: (a) 0.00Pd-CuO, (b) 1.25Pd-CuO Fig. 6. (a) Responses of 0.00Pd-CuO, 0.75Pd-CuO, 1.25Pd-CuO and 1.50Pd-CuO sensors to 50 ppm H2S as a function of operating temperature, the insert is the magnification of response operated between 200 oC to 340 oC. (b) Response of the 1.25Pd-CuO to different relative humidity water vapor. Fig. 7. Gas responses (Ra/Rg or Rg/Ra) to 50 ppm of different gas at 80 oC with different Pd doping. Fig. 8. Sensing transients of 1.25Pd-CuO to 50 ppm H2S at 80 oC. Fig. 9. (a) (a) The response of 1.25Pd-CuO to 1 ppm H2S with different heating time during recovery process, the insert is the sensing diagram of the gas sensor with a 4.6V electric modulation voltage. (b) The recovery cycles for 1 ppm H2S of the 1.25Pd-CuO sensor at 80 oC with a 4.6V electricd modulation voltage Fig. 10. (a) Gas response of 1.25Pd-CuO sensor to different concentrations of H2S at 80 oC. (b) Long-term stability of 1.25Pd-CuO sensor to 50 ppm H2S at 80 oC. Fig. 11. A scheme to illustrate the sensing mechanism for Pd doped CuO.

20

Biographies Xiaobing Hu received his bachelor degree in Materials Chemistry from Shanghai Polytechnic University (SSPU) in 2015, and currently is continuing to pursue a master degree in Shanghai Polytechnic University. His research focuses on the chemical gas sensors. Zhigang Zhu received a Ph.D. degree in Shanghai Institute of Ceramics, Chinese Academy of Sciences (2005), where he worked on functional materials and devices. In 2009 - 2012, he was a Research Associate at the University of Cambridge working in MEMS design and fabrication for biosensors. Since March 2012, he is a Jinqiao Professor in Shanghai Polytechnic University. He has published over 40 peer-reviewed journal papers so far and his main research interests are Micro-/Nano- biosensors and chemical gas sensors. Cheng Chen obtained his BS degree from School of Materials Science and Engineering, East China University of Science and Technology (ECUST) in 2006. He received his PhD degree from Key Laboratory for Ultrafine Materials of Ministry of Education, ECUST in 2011, focused on photonic crystal and hydrogel sensor materials. He is an associate professor in Shanghai Polytechnic University and his current research interests include functionalization and fabrication of nanomaterials and devices for sensing and biomedical application. Tianyang Wen is majoring in materials chemistry at the Shanghai Polytechnic University, and will start his master course in 2017. His research focuses on the chemical gas sensors. Xueling Zhao received her Ph.D. in Analytical Chemistry from the East China University of Science and Technology (ECUST) in 2014, and currently is a lecturer in Shanghai Polytechnic University. Her research focuses on the multifunctional nanomaterials and electrochemical sensors. LiLi Xie received a Ph.D. degree in Shanghai Institute of Ceramics, Chinese Academy of Sciences (2005), where she worked on synthesis, characterization and properties of molecular sieves. She is an associate professor in Shanghai Polytechnic University, and has published over 15 peer-reviewed journal papers so far. Her main research interests are the gas sensing properties of inorganic nano-materials.

21

Figure1

Figure2

Figure3

Figure4

Figure5

Figure6

Figure7

Figure8

Figure9

Figure10

Figure11