Cerium-Doped Copper(II) Oxide Hollow Nanostructures as Efficient

6 hours ago - The work also shows that the use of hollow nanostructures could be the .... Chemical composition and the presence of both oxygen vacanci...
0 downloads 5 Views 3MB Size
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article Cite This: ACS Omega 2018, 3, 5029−5037

Cerium-Doped Copper(II) Oxide Hollow Nanostructures as Efficient and Tunable Sensors for Volatile Organic Compounds Inderjeet Singh,† Sayan Dey,‡ Sumita Santra,† Katharina Landfester,*,§ Rafael Muñoz-Espí,*,§,∥ and Amreesh Chandra*,† †

Department of Physics and ‡Department of Electronics and Electrical Communications, Indian Institute of Technology, Kharagpur 721302, West Bengal, India § Department of Physical Chemistry of Polymers, Max Planck Institute for Polymer Research, Ackermannweg 10, Mainz 55128, Germany ∥ Institute of Materials Science (ICMUV), University of Valencia, C/Catedràtic José Beltrán 2, Paterna 46980, Spain S Supporting Information *

ABSTRACT: Tuning sensing capabilities of simple to complex oxides for achieving enhanced sensitivity and selectivity toward the detection of toxic volatile organic compounds (VOCs) is extremely important and remains a challenge. In the present work, we report the synthesis of pristine and Ce-doped CuO hollow nanostructures, which have much higher VOC sensing and response characteristics than their solid analogues. Undoped CuO hollow nanostructures exhibit high response for sensing of acetone as compared to commercial CuO nanoparticles. As a result of doping with cerium, the material starts showing selectivity. CuO hollow structures doped with 5 at. % of Ce return highest response toward methanol sensing, whereas increasing the Ce doping concentration to 10%, the material shows high response for bothacetone and methanol. The observed tunability in selectivity is directly linked to the varying concentration of the oxygen defects on the surface of the nanostructures. The work also shows that the use of hollow nanostructures could be the way forward for obtaining high-performance sensors even by using conventional and simple metal or semiconductor oxides.

1. INTRODUCTION With the stringent implementation of rules, which ensure human safety from air pollution in industries, ranging from cottage to defense, the development of efficient gas sensors has gained much attention.1 Also in this case, the dangers associated with sustained exposure to chemical vapors remain sometimes underestimated. Volatile organic compounds (VOCs, e.g., acetone, methanol, ethanol, etc.) possess high vapor pressures at room temperature, which increases their exposure to humans, causing adverse health effects.2 For example, acetone, an extensively used chemical in research laboratories and industry, is easily combustible, evaporates at room temperature, and can be medically dangerous, if inhaled for sustained periods. Unfortunately, these aspects are ignored by many users. Another common organic solvent routinely used is methanol. It also attracts high attention, as it is produced as a primary component in biodiesel production and is a needed additive for dyes, etc.3−5 A variety of gas sensors have been developed based on different operating principles, such as resistive (oxide or organic semiconductors), electrochemical, surface acoustic wave, and field effect types.6−9 Among these gas sensors, semiconducting © 2018 American Chemical Society

metal oxide-based resistive gas sensors are most widely investigated owing to their low cost, ease of production, simple measurement, higher sensitivity, lower detection limits, versatility, and fast recovery times. Most commercial products use metal oxides based on n- or p-type semiconductors, such as SnO2, WO3, ZnO, CuO, Co3O4, and NiO, for the detection of toxic VOCs.10−15 Surface modification of these metal oxides such as ZnO by noble metals, such as Au and Ag, has also been used as a strategy for enhanced sensing toward acetone, ethanol, and trimethylamine.16−18 Composites of metal oxide viz. SnO2, Au, and graphene oxide have also been used as an alternative for enhanced ethanol sensing.2 However, the performance of these metal oxides is still limited because of their low selectivity toward different VOCs. To improve the performance and to induce higher surface kinetics, the strategy of altering the particle morphology and defect concentrations by using suitable dopants has been proposed.19,20 However, this doping leads to an appreciable increase in the cost. It must be Received: February 7, 2018 Accepted: April 2, 2018 Published: May 9, 2018 5029

DOI: 10.1021/acsomega.8b00203 ACS Omega 2018, 3, 5029−5037

Article

ACS Omega

Scheme 1. Scheme for the Synthesis of Ce-Doped CuO Hollow Nanostructures Using the Inverse Miniemulsion Strategy

2.2. Structural, Morphological, and Chemical Characterization. For structural characterization of the prepared samples, powder X-ray diffraction patterns were recorded using a PANalytical high-resolution diffractometer PW 3040/60 with Cu Kα radiation (λ = 1.54 Å) as the incident wavelength. The diffraction pattern was collected in the 2θ range 10°−100°, and structure refinement was performed using the Fullprof software. The morphology of the samples was investigated using transmission electron microscopy (TEM) with a JEOL 1400 microscope. For TEM measurements, small amounts of the colloidal dispersions were diluted in an identical organic phase and drop-casted onto carbon-coated Cu grids. 2.3. Gas Sensing Measurements. To perform gas sensing experiments, the hollow nanostructured CuO-based sensing materials were dispersed in 20 mL of acetone and subsequently drop-casted on a Pt-based interdigitized electrode (Synkera Technologies) with an interfinger gap of 100 μm. The device was then mounted on a hot chuck fitted in a stainless steel airtight probe station and probed for the subsequent measurements. The gas flow was controlled by digital mass flow controllers (Alicat Instruments Ltd.), and the data were recorded using Agilent 34972A LXI data acquisition card for further analysis. The baseline resistance was confirmed by flowing compressed air at a rate of 500 sccm for 3 h. The VOC (in the liquid form) was taken in a bubbler, and air was bubbled to form subsequent vapors before being passed into the chamber. In this work, the sensor response was defined as the ratio Rgas/Rair, where Rair and Rgas are the resistance of the sensing material in the presence of air and different analyte VOCs, respectively.

mentioned that many novel and intriguing morphologies have been synthesized and used.21,22 Various morphologies of CuO and derived materials have been earlier investigated for sensing NO2, H2S, NH3, CO, and C2H5OH in a working temperature range of 200−400 °C.23−27 Recently, its application has also been shown for detection of different VOCs.15 In this work, using this conventionally used simple oxide of copper (i.e., CuO), it is shown that a hitherto unexplored strategy of using hollow nanostructures can bring quantum jump in gas-sensing capabilities. The results, after comparison with the available literature, clearly suggest that using composites based on expensive components, such as graphene and reduced graphite oxide, may not actually be required if the hollow particles are carefully tuned. These results would lead to a significant cost reduction of the final device.

2. METHODS 2.1. Synthesis of Ce-Doped CuO Nanostructures. For the synthesis of undoped and Ce-doped CuO hollow nanostructures, the technique of inverse miniemulsion was followed.28,29 In a typical synthesis procedure, as shown in Scheme 1, the dispersed phase was prepared by dissolving stoichiometric amounts of precursor salts Ce(NO3)3·6H2O (Fluka, p.a., ≥99.0%) and Cu(NO3)2·2H2O (Sigma-Aldrich, ≥99%) in water while maintaining a concentration of 0.5 M. A solution (1 wt %) of the surfactant poly(isobutylene succinimide pentamine) (PIBSP, OS85737, Lubrizol France, Rouen) in toluene (Sigma-Aldrich, 99.7%) was prepared and used as the continuous phase. The dispersed and continuous phases were mixed in a 1:4 weight ratio and pre-emulsified by stirring for 30 min at 1000 rpm. Subsequently, the mixture was ultrasonified for 3 min (Branson Sonifier W-450D, 1/2″ tip, 70% amplitude, 1 s pulse, 0.1 s pause) while it was kept in an ice bath. Triethylamine (TEA, Sigma-Aldrich, ≥99.5%), in a proportion of 3 equiv with respect to the Cu(II) precursor, was rapidly added by using a syringe. The emulsion was kept under constant stirring for 4 h at 1000 rpm at 80 °C to ensure homogenization and completion of the reaction. The precipitate was separated from the dispersion by centrifugation for 20 min at 4000 rpm. The obtained powders were washed with ethanol (twice), deionized water (twice), and acetone (once). After each washing process, the dispersion was centrifuged for 10 min at 4000 rpm. To investigate the gassensing properties, dense undoped and Ce3+-doped CuO powders were obtained by a thermal treatment at 250 °C for 4 h.

3. RESULTS AND DISCUSSION 3.1. Crystal Structure and Morphology. The minimization of agglomeration during the growth of nanoparticles is a critical issue to make the synthetic protocols at low temperatures more useful. As mentioned in Methods, in the present case, the crystallization of the hollow nanostructures was induced at a reasonably low temperature of 80 °C. For most metal oxides, single-phase formation at such low temperatures is difficult. Therefore, X-ray diffraction studies were performed to prove that single-phase CuO and Ce-doped CuO can be achieved, even at such low to moderate temperatures. Powder X-ray diffraction patterns for the samples under investigation are shown in Figure 1. The diffraction profiles are consistent with those expected for polycrystalline samples. The peaks can be indexed using the monoclinic phase of CuO (JCPDS card no. 48-1548), with no discernible signatures due to impurity phase(s). 5030

DOI: 10.1021/acsomega.8b00203 ACS Omega 2018, 3, 5029−5037

Article

ACS Omega

reactants at the interface. The growth of Ce-doped CuO hollow structures can be rationalized by the simplified mechanism depicted in Figure 3. Because of the presence of the base TEA and the surfactant PIBSP, metallic Cu ions precipitate in the form of Cu(OH)2 nanorods at the droplet interface, resulting in the formation of hollow nanostructures made up by the entanglement of nanorods at the exterior. However, along with these conditions, when Ce ions are added to the system to prepare the doped samples, Ce ions also precipitate as particulate Ce(OH)3, formed along with the precipitated Cu(OH)2 nanorods (Figure 3a,b). This induces the formation of Ce-doped hollow nanostructures with both nanorods and nanoparticles at the exterior (Figure 3c). Smaller precipitated particles of cerium hydroxide possess higher solubility in the solution, and therefore there is a higher probability for Ostwald ripening to occur at the droplet interface, which can be a reason for the observable opening of the hollow morphologies for Cedoped CuO as compared to the undoped CuO. This would also explain the conversion toward better solid shells in the higher Ce-doped CuO hollow structures. 3.2. Chemical Characterization. Chemical composition and the presence of both oxygen vacancies and valence state of Ce in the samples were characterized by the X-ray photoelectron spectroscopy (XPS) measurements in the doped samples (Figure 4). As reported previously, undoped hollow nanostructures of CuO show only the presence of the oxidation state 2+ for Cu in the samples; hence, no oxygen vacant sites are found in the samples.28 The higher binding energy peak in the O 1s spectra refers to the oxygen absorbed on the surface. In the cases of the samples doped with 5 and 10% Ce, also the presence of the 2+ oxidation state of Cu was detected (Supporting Information Figure S3). The Ce 3d spectra shown in Figure 4a,b confirmed both 3+ and 4+ oxidation states of Ce in the doped samples. These spectra were deconvoluted similarly to previously reported, in the case of CeO2 nanostructures, for quantification of the 3+ and 4+ oxidation states (Tables S1 and S2 show the binding energies of the deconvoluted peaks, which are consistent with the reported data).33,34 Relative concentrations of the 3+ oxidation state of Ce in the samples were calculated to be 0.48 and 0.36, in the cases of 5 and 10% Ce-doped CuO, respectively. A similar behavior was observed in the case of oxygen defect concentration, which was confirmed by the deconvolution of O 1s spectra shown in Figure 4c,d. The higher binding energy peak of the O 1s spectra is attributed to the oxygen defects on the surface caused by the doping of Ce3+ and Ce4+ in CuO, which can be clearly seen from the deconvoluted peaks of the O 1s spectra. The relative concentration of the higher binding energy peak to the lower binding energy peak (OHBE/OLBE) ascribed to oxygen defects was higher for the 5% Ce-doped

Figure 1. XRD patterns of the undoped and 5 and 10% Ce-doped CuO.

The chemical or sensing capability in nanostructures is directly linked to the exposed surface, which can synergistically contribute in the overall electron-exchange kinetics. TEM micrographs of the CuO-based nanostructures, with up to 10% Ce doping, are shown in Figure 2a−c (Figure S1 shows the high-resolution TEM images of CuO hollow nanostructures). Both doped and undoped samples indicated the formation of hollow nanostructures, with varying cavity size. Additionally, the Ce-doped CuO hollow structures were predominantly made up by assembling of nanorods and nanoparticles. In comparison, the hollow nanostructures for the pure CuO (Figure 2a) were a convoluted picture of entangled rodlike structures appearing at the droplet interface. With increasing cerium concentrations, the particle morphology showed transformation from porous to solid shell-like. The results for samples with up to 20% cerium doping also showed a similar trend. The details are given in the Supporting Information (Figure S2). Formation of hollow structures within the confined droplets of miniemulsions can be achieved by enforcing precipitation at the droplet interface.30,31 These conditions can be produced by the combined effects of controlled diffusion of precipitating agent and increased concentration of reactant metallic ions (by complexation with the surfactant) near the droplet interface.28,32 In the present study, the amphiphilic nature of TEA and the electronegative amino groups from the surfactant PIBSP are important factors for the increased concentration of

Figure 2. TEM micrographs of (a) undoped and (b) 5 and (c) 10% Ce-doped CuO hollow nanostructures. 5031

DOI: 10.1021/acsomega.8b00203 ACS Omega 2018, 3, 5029−5037

Article

ACS Omega

Figure 3. Proposed schematic mechanism for the formation of Ce-doped CuO hollow nanostructures showing (a) complexation between surfactant amino groups and Cu ions, (b) precipitation using TEA, and (c) formation of the hollow nanostructures.

Figure 4. Deconvoluted XPS spectra of (a,c) Ce 3d and (b,d) O 1s core levels for 5 and 10% Ce-doped CuO samples.

found to be maximum at 200 °C (shown in Figure 5a). Subsequently, the sensing response at 200 °C in the concentration range 3500−14 000 ppm was tested. The related graphs are presented in Figure 5b. The sensor response varied from 22 to 32 times. Commercially available copper(II) oxide was also tested in the presence of acetone at 200 °C (also shown in Figure 5b). Transient gas response for all measurements is also plotted in actual sensor resistance values and shown in the Supporting Information (Figures S4 and S5). As evident from Figure 5b, the hollow nanostructures were highly sensitive as compared to their commercial counterparts. Such a variation may be attributed to the difference in the specific surface area and porosity of the two materials. The hollow nanostructures possess a higher surface area (∼85 m2/g, reported previously28) and porosity, which means an increased number of active adsorption sites for chemical sorption−desorption processes.

samples than 10% Ce-doped CuO. These (OHBE/OLBE) ratios were calculated to be 0.40 and 0.32 for the 5 and 10% Cedoped samples, respectively. The observed change in the Ce3+ and oxygen defect concentration can be explained by considering the doping concentration. At a lower doping concentration of Ce in CuO, the quantity of substituted Ce3+ ions will be higher and therefore the lower the probability of Ce3+ getting oxidized. At a higher doping concentration (say, 10%), Ce will reduce the quantity of substituted Ce ions and many Ce3+ clusters will exist in the samples, which can oxidize after precipitation. This will lead to reduced Ce3+ concentration as well as oxygen defects in the nanostructures. 3.3. Sensing Performance toward VOCs. The sensors show the highest response at an optimum temperature because of the preferential adsorption of O−, which is most reactive toward VOCs. Therefore, the temperature profile of the sensor (using 3500 ppm of acetone) was tested, and the response was 5032

DOI: 10.1021/acsomega.8b00203 ACS Omega 2018, 3, 5029−5037

Article

ACS Omega

cause a change in the optimum temperature. Therefore, the temperature profiles for the doped samples were also tested in the presence of 3500 ppm concentration of methanol and acetone. For both of these samples, the maximum operating temperature was found to be at 250 °C for both the VOCs (depicted in Figure 6a). The dynamic responses of 5 and 10%

Figure 5. Comparison of the sensing performance of commercial CuO and the CuO hollow nanostructures prepared in this work. (a) Sensitivity of the gas sensor to 3500 ppm acetone, over a temperature range of 150−300 °C, (b) transient gas response of the sensor with exposure to different concentrations of acetone at 200 °C, and (c) response values for different VOCs.

Changes in the surface area and porosity of Ce-doped CuO hollow nanostructures may affect the sensing properties.35 5 and 10% Ce-doped CuO hollow nanostructures showed higher specific surface areas viz., ∼144 and ∼154 m2/g, respectively, and smaller pores as compared to the undoped CuO hollow nanostructures (Figure S6). The CuO hollow nanostructure sensor response was also superior as compared to previous reports on CuO-based nanomaterials (Supporting Information Table S3). From the plot, the limit of detection (LOD) was also estimated for both the samples (Supporting Information). The LOD for the commercial sample was ∼3298 ppm, whereas it was ∼980 ppb for our CuO spheres. The response of the hollow nanostructures of CuO was also tested in the presence of several other VOCs including methanol, ethanol, chloroform, and toluene (presented in Figure 1c). It was found that the sensor had much higher response toward acetone as compared to other VOCs, indicating a high selectivity toward acetone. The sensing performance of 5 and 10% Ce-doped CuO hollow nanostructures was tested in the presence of different VOCs. However, a change in the chemical composition can

Figure 6. Sensing performance of 5 and 10% Ce-doped CuO samples: tunability of selectivity from methanol (for 5%) to acetone (for 10%). (a) Sensitivity of the gas sensor to 3500 ppm acetone, over a temperature range of 150−300 °C, and transient gas response of the sensor with exposure to different concentrations of (b) acetone and (c) methanol for 5 and 10% Ce-doped CuO at 250 °C.

Ce-doped CuO samples are presented in Figure 6b,c (from 900 to 14 500 ppm concentrations). As the doping percentage was increased from 0 to 5% Ce, the doped sample showed the highest response for methanol in comparison to other tested VOCs. The response for 5% Ce-doped CuO samples was 28 times toward methanol and 12 times for acetone, as evident from Figure 6. The response was found to be much lesser for other VOCs. As the doping percentage of Ce was increased from 5 to 10% in CuO samples, the sensor was more responsive toward both acetone and methanol. In this case (10% doped sample), the response was 22 times toward acetone and 18 times for 5033

DOI: 10.1021/acsomega.8b00203 ACS Omega 2018, 3, 5029−5037

Article

ACS Omega C7H8 + 4O− ↔ H 2O + C7H6 + 2O− + 2e−

methanol at 3500 ppm concentrations. The LOD values calculated for the 5% Ce-doped CuO samples for acetone and methanol were 9.80 and 1.03 ppm, respectively. For the 10% Ce-doped CuO samples, the calculated LOD values in the presence of acetone and methanol were 2.65 and 5.99 ppm, respectively. A comparison of the response of undoped and Cedoped CuO samples toward the tested VOCs is depicted in Figure 7. It can be clearly seen that the highest response toward

2CHCl3 + 2O− → 2COCl 2 + 2HCl + 2e− CH3OH + O− → HCOH + H 2O + e− C3H6O + 8O− → 3CO2 + 3H 2O + 8e−

In the case of ethanol, the reaction with adsorbed oxygen species depends upon the acid−base properties of the sensing material. Ethanol is decomposed to the intermediate state C2H4 in the case of acidic oxide, whereas it will convert to CH3CHO for basic oxide by the following reaction: C2H5OH → C2H4 + H 2O (acidic oxide) C2H5OH → CH3CHO + H 2 (basic oxide)

Copper oxide is basic in nature. Thus, it will dehydrate to CH3CHO. This intermediate state will react with surfaceadsorbed oxygen species, which is given below. CH3CHO + 5O− → 2CO2 + 2H 2O + 5e−

In all above reactions, electrons are released and lead to reduction of the hole concentration in p-type CuO-sensing material. As a result, the resistance of the sensing layer increases. It is also observed that the highest numbers of electrons are released in the case of acetone compared to other VOCs. The experimental results showed that only CuO has the highest response toward acetone at 200 °C, which suggests its maximum interaction with the surface oxygen species. When Ce is doped in CuO, Ce4+ is replaced with Cu2+. Thus, Ce stays at CeO2 form in copper oxide. Cerium oxide can show both basic and acidic properties. Thus, for cerium-doped copper oxide, both dehydrogenation to C2H4 and dehydration to CH3CHO reactions can happen. C2H4 will react with oxygen by the following equation.

Figure 7. Comparison of sensing performance of undoped and 5 and 10% Ce-doped CuO samples at a 3500 ppm concentration of VOCs.

acetone in the undoped CuO samples changed to highest response and selectivity to methanol in the 5% doped samples. In addition, the 10%-doped samples showed enhanced response for both acetone and methanol. Therefore, the doping of Ce in CuO samples causes tunability in the optimum response and selectivity toward VOCs, which depends predominantly on the doping concentrations of cerium. To understand the observed response in CuO hollow nanostructures toward acetone and changes after Ce doping, it is necessary to explain the mechanism for sensing of VOCs using metal oxide as active materials. When air comes in contact with the metal oxide surface, oxygen molecules from the air get adsorbed in the form of different species O−, O2−, and O2− onto the surface of metal oxides by trapping the free surface electrons from the conduction band of the metal oxides. The ionosorption of the species of O2−, O−, and O2− is known to be dominant at 400 °C, respectively.36 This adsorption process leads to the creation of more holes available near the surface and formation of a hole accumulation layer (HAL) in p-type metal oxides such as CuO. The adsorbed oxygen acts as active reaction sites for the reducing VOCs to get attached. Upon exposure of the VOCs, they are oxidized by these oxygen reactive species, resulting in the creation of electrons and decreasing the concentration of holes in the surface layer. Consequently, an increase in the resistance is observed in the p-type metal oxide semiconductors upon exposure to VOCs. The reactions mentioned in the mechanism are mentioned below: Adsorption of oxygen species:

C2H4 + 6O− → 2CO2 + 2H 2O + 6e−

According to the above discussion, it can be suggested that the adsorption of the oxygen reactive species on the surface of the metal oxide surface will determine the sensing performance toward different VOCs. This adsorption can be affected by several of the following reasons: (i) Different metal ions: this can cause the change in the electrical properties (total charge density) of the materials, which would cause the interaction between metal surface and adsorbing species to change, therefore changing the final adsorption. (ii) Temperature: it causes the selective adsorption of oxygen ions on the material surface. In the present case, the effects of the temperature were neutralized by operating the sensors at the optimum temperatures. (iii) Surface defects and roughness: defects such as rough surface, oxygen defects (vacancies and interstitial), and disordered structures and boundaries on the surface of materials can enhance the adsorption of oxygen from the atmosphere, which can improve the gas-sensing properties. (iv) Surface area and porosity: enhanced surface area and porosity can also enhance the oxygen ion adsorption and the properties. In addition to these, kinetic diameter of the adsorbing VOCs may also affect the sensing response.

O2ads + e− ↔ O2ads−

O2ads− + e− ↔ 2Oads−

Oads− + e− ↔ O2ads− When these materials are exposed to toluene, chloroform, methanol, or acetone, the following reactions may take place.37,38 5034

DOI: 10.1021/acsomega.8b00203 ACS Omega 2018, 3, 5029−5037

Article

ACS Omega

reported enhanced response with Nd doping in the case of alcohol (ethanol) as compared to the undoped samples where it decreases for other VOCs (methanol). Even, the trend of the response with doping was not found similar for different VOCs. In that work, the authors predominantly discussed the case of increase in the sensing toward ethanol. However, the cases where the sensing was not enhanced or contradictory were not debated. In our case, data analysis of the doped nanostructures confirms that the sensing of different VOCs shows a possible relation/trend with the concentration of the surface defects in the materials. We observe that the increase in oxygen defect concentration works in favor for the sensing of methanol, whereas this is opposite in the case of acetone. In addition to the reasons mentioned above, the difference in the sensitivity of different VOCs with respect to the defect concentration could also be due to the physical properties of VOCs. As the kinetic diameter of methanol (0.36 nm) is lower than that of acetone (∼0.469 nm), methanol could be more adsorbing at the defect surfaces.38,40 In addition, oxygen vacant sites in the crystal structure refer to deficiency of negative charge; therefore, metal oxide can be considered as showing acidic properties.41 A difference in the electronegative character of the VOCs due to the presence of different electronegative groups could cause the interaction of the molecules with the adsorbed oxygen or the metal oxide surface to change and can make it easier or difficult for metal oxides to accept the electrons. This could also be responsible for the observed change in the sensing behavior for the samples with varying defect concentrations.

The effects of the metal ion doping on the surface charge were characterized by the zeta potential measurements. From the measurement data shown in Figure 8, the surface negativity

Figure 8. Zeta potential of 5 and 10% Ce-doped CuO hollow nanostructure dispersions in water at 50 mg/L concentration and pH 7.0.

was found to be maximum for CuO. The surface negativity decreases, as we increase the Ce amount, which is obvious because Cu2+ is replaced by Ce3+. Hence, as the number of electrons increase, the value of HAL decreases. This leads to reduced surface negativity and resistance of the sensing material. This can be a possible reason for the enhanced sensitivity. However, as expected, with higher doping, the negativity also decreases; therefore, the decrease in methanol response and the increase in acetone response at 10% Ce doping as compared to 5% Ce doping cannot be explained. The changes observed in the surface area and porosity due to the doping and change in morphology can cause the increase in the performance toward sensing, but it is also difficult to explain why this change will be positive for one VOC and not for the other ones tested. Defects on the surface, such as oxygen defects, can also alter the performance. A higher number of defects on the surface can create more sites for oxygen adsorption, which leads to a higher concentration of oxygen reactive species at the surface. Therefore, a high number of reducing gas molecules can react with the oxygen species at the surface, leading to higher change in the resistance. Therefore, a higher response would be observed at doping a concentration of Ce3+ in CuO, where the concentration of defects would be highest. The relative concentration of the oxygen defects, calculated by the analysis of the XPS data, is found to be 0.40 in the case of 5% doping of Ce3+, which was much higher than the value of 0.32 in the 10% doping of Ce3+ in CuO. This is compared to the undoped CuO samples, where the defects would be minimal. If we compare this and the data shown in Figure 7, it is clear that the sensing of methanol clearly correlates with the concentration of oxygen defects. At the highest oxygen defect concentration in the 5%doped samples, the maximum response toward methanol is observed, which is lower in 10% Ce-doped and undoped CuO samples. Similarly, in the case of acetone, the highest response is observed when the defect concentrations are lower, that is, in the undoped samples followed by 10 and 5% Ce-doped CuO samples. Therefore, by careful analysis, we can say that in our samples, oxygen defects can cause the sensing response toward different VOCs to change. Similar changes in the VOC-sensing performance have been earlier reported in Nd-doped SnO2 materials.39 The authors

4. CONCLUSIONS Undoped and Ce-doped CuO hollow nanostructures synthesized using the inverse miniemulsion strategy show enhancement and tuning in the sensing response toward VOCs. Changes observed in the morphological nature of the Ce-doped hollow nanostructures are explained using the growth mechanism in confined droplets. Hollow nanostructures of undoped CuO show extremely high response and selectivity toward acetone as compared to the commercial CuO. In the case of 5% Ce-doped hollow structures, high response and selectivity for methanol is observed, and at 10% Ce doping, high response and selectivity toward both acetone and methanol is observed simultaneously. Amongst the various possible reasons for the observed increase in response and tunability in the VOC selectivity, the change in the oxygen defect concentration upon Ce doping is found to be a possible and plausible reason as compared to other reasons such as metal ion dependency, temperature, and surface area. The observed enhancement in the sensing response of CuO hollow nanostructures makes it extremely important to study other hollow metal oxide-based nanomaterials for sensing applications.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsomega.8b00203. Morphological characterization, calculation details for LOD, XPS measurements, transient gas response in actual resistance values, surface area characterization, data for deconvolution of Ce 3d and O 1s XPS spectra, and comparison of sensing response with previous reports (PDF) 5035

DOI: 10.1021/acsomega.8b00203 ACS Omega 2018, 3, 5029−5037

Article

ACS Omega



(13) Nguyen, H.; El-Safty, S. A. Meso- and Macroporous Co3O4 Nanorods for Effective VOC Gas Sensors. J. Phys. Chem. C 2011, 115, 8466−8474. (14) Gao, Y.; Kong, Q.; Zhang, J.; Xi, G. General Fabrication and Enhanced VOC Gas-Sensing Properties of Hierarchically Porous Metal Oxides. RSC Adv. 2017, 7, 35897−35904. (15) Xia, S.; Zhu, H.; Cai, H.; Zhang, J.; Yu, J.; Tang, Z. Hydrothermally Synthesized CuO based Volatile Organic Compound Gas Sensor. RSC Adv. 2014, 4, 57975−57982. (16) Meng, F.; Hou, N.; Jin, Z.; Sun, B.; Li, W.; Xiao, X.; Wang, C.; Li, M.; Liu, J. Sub-ppb detection of acetone using Au-modified flowerlike hierarchical ZnO structures. Sens. Actuators, B 2015, 219, 209− 217. (17) Meng, F.; Hou, N.; Jin, Z.; Sun, B.; Guo, Z.; Kong, L.; Xiao, X.; Wu, H.; Li, M.; Liu, J. Ag-decorated ultra-thin porous single-crystalline ZnO nanosheets prepared by sunlight induced solvent reduction and their highly sensitive detection of ethanol. Sens. Actuators, B 2015, 209, 975−982. (18) Meng, F.; Zheng, H.; Sun, Y.; Li, M.; Liu, J. Trimethylamine Sensors Based on Au-Modified Hierarchical Porous Single-Crystalline ZnO Nanosheets. Sensors 2017, 17, 1478. (19) Zhao, Q.; Ju, D.; Deng, X.; Huang, J.; Cao, B.; Xu, X. Morphology-Modulation of SnO2 Hierarchical Architectures by Zn Doping for Glycol Gas Sensing and Photocatalytic Applications. Sci. Rep. 2015, 5, 7874. (20) Yogamalar, R.; Mahendran, V.; Srinivasan, R.; Beitollahi, A.; Kumar, R. P.; Bose, A. C.; Vinu, A. Gas-Sensing Properties of NeedleShaped Ni-Doped SnO2 Nanocrystals Prepared by a Simple Sol-Gel Chemical Precipitation Method. Chem.−Asian J. 2010, 5, 2379−2385. (21) Ma, X.; Zhou, X.; Gong, Y.; Han, N.; Liu, H.; Chen, Y. MOFderived Hierarchical ZnO/ZnFe2O4 Hollow Cubes for Enhanced Acetone Gas-Sensing Performance. RSC Adv. 2017, 7, 34609−34617. (22) Wang, C.; Cheng, X.; Zhou, X.; Sun, P.; Hu, X.; Shimanoe, K.; Lu, G.; Yamazoe, N. Hierarchical alpha-Fe2O3/NiO Composites with a Hollow Structure for a Gas Sensor. ACS Appl. Mater. Interfaces 2014, 6, 12031−12037. (23) Liu, X.; Hu, M.; Wang, Y.; Liu, J.; Qin, Y. High Sensitivity NO2 Sensor based on CuO/p-Porous Silicon Heterojunction at Room Temperature. J. Alloys Compd. 2016, 685, 364−369. (24) Vuong, N. M.; Chinh, N. D.; Huy, B. T.; Lee, Y.-I. CuODecorated ZnO Hierarchical Nanostructures as Efficient and Established Sensing Materials for H2S Gas Sensors. Sci. Rep. 2016, 6, 26736. (25) Bhuvaneshwari, S.; Gopalakrishnan, N. Hydrothermally Synthesized Copper Oxide (CuO) Superstructures for Ammonia Sensing. J. Colloid Interface Sci. 2016, 480, 76−84. (26) Steinhauer, S.; Singh, V.; Cassidy, C.; Gspan, C.; Grogger, W.; Sowwan, M.; Köck, A. Single CuO nanowires Decorated with SizeSelected Pd Nanoparticles for CO Sensing in Humid Atmosphere. Nanotechnology 2015, 26, 175502. (27) Wang, F.; Li, H.; Yuan, Z.; Sun, Y.; Chang, F.; Deng, H.; Xie, L.; Li, H. A Highly Sensitive Gas sensor based on CuO Nanoparticles Synthetized Via a Sol−Gel Method. RSC Adv. 2016, 6, 79343−79349. (28) Singh, I.; Landfester, K.; Chandra, A.; Muñoz-Espí, R. A New Approach for Crystallization of Copper(II) Oxide Hollow Nanostructures with Superior Catalytic and Magnetic Response. Nanoscale 2015, 7, 19250−19258. (29) Antonello, A.; Jakob, G.; Dolcet, P.; Momper, R.; Kokkinopoulou, M.; Landfester, K.; Muñoz-Espí, R.; Gross, S. Synergy of Miniemulsion and Solvothermal Conditions for the Low-Temperature Crystallization of Magnetic Nanostructured Transition-Metal Ferrites. Chem. Mater. 2017, 29, 985−997. (30) Muñoz-Espí, R.; Mastai, Y.; Gross, S.; Landfester, K. Colloidal Systems for Crystallization Processes from Liquid Phase. CrystEngComm 2013, 15, 2175−2191. (31) Muñoz-Espí, R.; Weiss, C. K.; Landfester, K. Inorganic Nanoparticles Prepared in Miniemulsion. Curr. Opin. Colloid Interface Sci. 2012, 17, 212−224.

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (K.L.). *E-mail: [email protected] (R.M.-E.). *E-mail: [email protected] (A.C.). ORCID

Katharina Landfester: 0000-0001-9591-4638 Rafael Muñoz-Espí: 0000-0002-8146-2332 Amreesh Chandra: 0000-0001-8247-7926 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors acknowledge the financial support by IGSTC (DST, India) and the Max Planck Society (Germany) for funding under their program of Max Planck Partner Group on “Multifunctional Hybrid Nanostructures for Alternative Energy Systems” headed by Prof. A.C. at IIT Kharagpur, India.



REFERENCES

(1) Polleux, J.; Gurlo, A.; Barsan, N.; Weimar, U.; Antonietti, M.; Niederberger, M. Template-Free Synthesis and Assembly of SingleCrystalline Tungsten Oxide Nanowires and their Gas-Sensing Properties. Angew. Chem. 2006, 118, 267−271. (2) Meng, F.; Zheng, H.; Chang, Y.; Zhao, Y.; Li, M.; Wang, C.; Sun, Y.; Liu, J. One-Step Synthesis of Au/SnO2/RGO Nanocomposites and Their VOC Sensing Properties. IEEE Trans. Nanotechnol. 2018, 17, 212−219. (3) Chang, Y.; Tang, N.; Qu, H.; Liu, J.; Zhang, D.; Zhang, H.; Pang, W.; Duan, X. Detection of Volatile Organic Compounds by Selfassembled Monolayer Coated Sensor Array with Concentrationindependent Fingerprints. Sci. Rep. 2016, 6, 23970. (4) Wang, C.; Dong, R.; Wang, X.; Lian, A.; Chi, C.; Ke, C.; Guo, L.; Liu, S.; Zhao, W.; Xu, G.; Li, E. Exhaled Volatile Organic Compounds as Lung Cancer Biomarkers During One-Lung Ventilation. Sci. Rep. 2014, 4, 7312. (5) Peng, G.; Tisch, U.; Adams, O.; Hakim, M.; Shehada, N.; Broza, Y. Y.; Billan, S.; Abdah-Bortnyak, R.; Kuten, A.; Haick, H. Diagnosing Lung Cancer in Exhaled Breath using Gold Nanoparticles. Nat. Nanotechnol. 2009, 4, 669−673. (6) Liao, F.; Chen, C.; Subramanian, V. Organic TFTs as Gas sensors for Electronic Nose Applications. Sens. Actuators, B 2005, 107, 849− 855. (7) Miasik, J. J.; Hooper, A.; Tofield, B. C. Conducting Polymer Gas Sensors. J. Chem. Soc., Faraday Trans. 1986, 82, 1117−1126. (8) Peng, N.; Zhang, Q.; Lee, Y. C.; Tan, O. K.; Marzari, N. Gate Modulation in Carbon Nanotube Field Effect Transistors-based NH3 Gas Sensors. Sens. Actuators, B 2008, 132, 191−195. (9) Penza, M.; Aversa, P.; Cassano, G.; Wlodarski, W.; Kalantarzadeh, K. Layered SAW Gas Sensor with Single-Walled Carbon Nanotubebased Nanocomposite Coating. Sens. Actuators, B 2007, 127, 168−178. (10) Lee, D.-S.; Jung, J.-K.; Lim, J.-W.; Huh, J.-S.; Lee, D.-D. Recognition of Volatile Organic Compounds using SnO2 Sensor Array and Pattern Recognition Analysis. Sens. Actuators, B 2001, 77, 228− 236. (11) Bao, D.; Gao, P.; Wang, L.; Wang, Y.; Chen, Y.; Chen, G.; Li, G.; Chang, C.; Qin, W. ZnO Nanorod Arrays and Hollow Spheres through a Facile Room-Temperature Solution Route and Their Enhanced Ethanol Gas-Sensing Properties. ChemPlusChem 2013, 78, 1266−1272. (12) Perfecto, T. M.; Zito, C. A.; Volanti, D. P. Room-Temperature Volatile Organic Compounds Sensing based on WO3·0.33H2O, Hexagonal-WO3, and their Reduced Graphene Oxide Composites. RSC Adv. 2016, 6, 105171−105179. 5036

DOI: 10.1021/acsomega.8b00203 ACS Omega 2018, 3, 5029−5037

Article

ACS Omega (32) Singh, I.; Nigam, A. K.; Landfester, K.; Muñoz-Espí, R.; Chandra, A. Anomalous Magnetic Behavior Below 10 K in YCrO3 Nanoparticles Obtained under Droplet Confinement. Appl. Phys. Lett. 2013, 103, 182902. (33) Singh, I.; Landfester, K.; Muñoz-Espí, R.; Chandra, A. Evolution of Hollow Nanostructures in Hybrid Ce1‑xCuxO2 under Droplet Confinement Leading to Synergetic Effects on the Physical Properties. Nanotechnology 2017, 28, 075601. (34) Singh, I.; Chandra, A. Use of the Oxygen Storage Material CeO2 as Co-Catalyst to Improve the Performance of Microbial Fuel Cells. Int. J. Hydrogen Energy 2016, 41, 1913−1920. (35) Li, H.; Meng, F.; Liu, J.; Sun, Y.; Jin, Z.; Kong, L.; Hu, Y.; Liu, J. Synthesis and gas sensing properties of hierarchical meso-macroporous SnO2 for detection of indoor air pollutants. Sens. Actuators, B 2012, 166−167, 519−525. (36) Santra, S.; De Luca, A.; Bhaumik, S.; Ali, S. Z.; Udrea, F.; Gardner, J. W.; Ray, S. K.; Guha, P. K. Dip Pen NanolithographyDeposited Zinc Oxide Nanorods on a CMOS MEMS Platform for Ethanol Sensing. RSC Adv. 2015, 5, 47609−47616. (37) Santra, S.; Sinha, A. K.; De Luca, A.; Ali, S. Z.; Udrea, F.; Guha, P. K.; Ray, S. K.; Gardner, J. W. Mask-Less Deposition of Au-SnO2 Nanocomposites on CMOS MEMS Platform for Ethanol Detection. Nanotechnology 2016, 27, 125502. (38) Rakshit, T.; Santra, S.; Manna, I.; Ray, S. K. Enhanced Sensitivity and Selectivity of Brush-Like SnO2 nanowire/ZnO Nanorod Heterostructure based Sensors for Volatile Organic Compounds. RSC Adv. 2014, 4, 36749−36756. (39) Qin, G.; Gao, F.; Jiang, Q.; Li, Y.; Liu, Y.; Luo, L.; Zhao, K.; Zhao, H. Well-Aligned Nd-doped SnO2 Nanorod Layered Arrays: Preparation, Characterization and Enhanced Alcohol-Gas Sensing Performance. Phys. Chem. Chem. Phys. 2016, 18, 5537−5549. (40) Nayak, A. K.; Ghosh, R.; Santra, S.; Guha, P. K.; Pradhan, D. Hierarchical Nanostructured WO3-SnO2 for Selective Sensing of Volatile Organic Compounds. Nanoscale 2015, 7, 12460−12473. (41) Horzum, N.; Tascioglu, D.; Ö zbek, C.; Okur, S.; Demir, M. M. VOC Sensors based on a Metal Oxide Nanofibrous Membrane/QCM System Prepared by Electrospinning. New J. Chem. 2014, 38, 5761− 5768.

5037

DOI: 10.1021/acsomega.8b00203 ACS Omega 2018, 3, 5029−5037