A novel nitrogen dioxide gas sensor based on TiO2-supported Au nanoparticles: a van der Waals corrected DFT study
- First Online:
- 398 Downloads
The interactions of nitrogen dioxide molecule with TiO2-supported Au nanoparticles were investigated using density functional theory. Surface Au atoms on the TiO2-supported Au overlayer were found to be the most favorable binding sites, thus making the adsorption process very strong. Both oxygen and nitrogen atoms of the NO2 molecule can bind to the Au surface by forming strong chemical bonds. The adsorption of NO2 molecule on the considered structures gives rise to significant changes in the bond lengths, bond angles, and adsorption energies of the complex systems. The results indicate that NO2 adsorption on the TiO2-supported Au nanoparticle by its oxygen atoms is energetically more favorable than the NO2 adsorption by its nitrogen atom, indicating the strong binding of NO2 to the TiO2-supported Au through its oxygen atoms. Thus, the bridge configuration of TiO2/Au + NO2 is found to be the most stable configuration. Both oxygen and nitrogen atoms of NO2 move favorably towards the Au surface, as confirmed by significant overlaps in the PDOSs of the atoms that forming chemical bonds. This study not only suggests a theoretical basis for gas-sensing properties of the TiO2-supported Au nanoparticles, but also offers a rational approach to develop nanostructure-based chemical sensors with improved performance.
KeywordsDensity functional theory NO2 TiO2-supported Au nanoparticle PDOS
TiO2 is one of the most broadly studied transition metal semiconductors with outstanding properties, such as non-toxicity, high catalytic efficiency, and extensive bandgap . Until now, various kinds of well-known applications have been proposed for TiO2, such as photo-catalysis, gas sensor devices, organic dye-sensitized solar cells, water splitting, and air pollution control [2, 3, 4, 5]. Anatase, rutile, and brookite are the most important polymorphs of TiO2 . Of the three polymorphs of TiO2, the rutile form is found to be the most stable phase. There is not any detailed theoretical investigation on the physical and chemical properties of brookite because of its metastable property. This meta-stability results in some troubles during the synthesis of brookite . The improved reactivity of anatase is comparable with that of rutile and brookite phases [8, 9, 10, 11, 12, 13, 14]. Anatase has been extensively studied due to its enhanced activity in some photo-catalysis reactions, such as TiO2-supported metal particle reactions, compared to the rutile and brookite phases [15, 16, 17]. Unfortunately, as a most promising material, the widely application of TiO2-based gas sensors is influenced by its wide bandgap (3–3.2 eV). This results in the absorption of a small percentage of the incoming solar light (3–5%). An enormous amount of effort has been invested in enhancing the optical response of TiO2 by nitrogen doping .
Recently, gold was considered as an inactive metal, which possesses less activity than the other metals in many reactions. Haruta and co-workers showed that gold particles can increase the combustion of CO molecule and promote different catalytic reactions . The gold particles supported by metal oxides (oxide-supported gold particles) have gained more attention due to their higher activities in the surface processes [19, 20, 21, 22]. This leads to the structures with enhanced catalytic activity and higher stability [23, 24]. There are a large number of important reactions, in which the oxide-supported Au overlayers play a key role, including the epoxidation of C3H6 , reduction of NOx molecules , and dissociation of SO2 molecule . TiO2 has been considered as one of the most appropriate support materials for gold particles [28, 29]. The interactions of gold nanoparticles with TiO2 (rutile and anatase) have been widely studied in the last few years. Vittadini et al. considered the adsorption behaviors of gold clusters on the TiO2 anatase (101) surfaces . Metiu and co-workers investigated the adsorption site and electronic structures of the TiO2 rutile-supported Au nanoparticles .
The consecutive adsorption of NO2 molecules on the TiO2-supported Au overlayers probably produces N2 molecule formed from the central nitrogen atoms of the two adsorbed NO2 molecules. This is a consequence of the formation of weak chemical bonds between oxygen atoms of NO2 molecule and Ti sites of the adsorbent. This leads to weakening of the bond between central nitrogen and the side oxygen atoms of the adsorbed NO2 molecules. Based on this fact, we can conclude desorption of NO2 molecule from the surface of the TiO2-supported Au overlayer. The next NO2 molecule then can be adsorbed on the considered nanocomposite, and this consecutive process was repeated over and over again to obtain the enhanced sensing performance of the adsorbent material. Figure 1 shows a schematic structure of a metal oxide-based gas sensor. We have also commented on the charge analysis of the complex system according to the Mulliken population analysis. In this study, the main objective is to perform a systematic investigation on the adsorption behaviors of the TiO2-supported Au nanoparticles as potentially efficient gas sensors for NO2 detection.
Details of computation
All of DFT calculations [34, 35] were carried out using the Open source Package for Material eXplorer (OPENMX3.8) . OPENMX is an efficient software package for nano-scale materials simulations based on norm-conserving pseudo-potentials and pseudo-atomic localized basis functions [37, 38]. To optimize the structures, the pseudoatomic orbitals (PAOs) centered on atomic sites were used as basis sets. The calculations were done with a considered energy cutoff of 150 Ry. Pseudo-atomic orbitals were constructed by minimal basis sets (three s-state, three p-state, and one d-state radial functions) for the Ti, (three s-state, three p-state, two d-state, and one f-state radial functions) for the Au, and (two s-state, two p-state radial functions) for O and N atoms, within cut-off radii of basis functions set to the values of seven for Ti, nine for Au, and five for O and N (all in Bohrs). The total energy of the system was computed within the Perdew–Burke–Ernzerhof (PBE) form of the generalized gradient approximation (GGA) exchange–correlation potential . Mulliken population analysis was also conducted to fully analyze the charge transfer between NO2 and nanocomposite. To optimize the adsorption configurations of the TiO2-supported Au overlayers with adsorbed NO2 molecules, all atoms of the system are entirely relaxed until the force on each atom is less than 0.01 eV/Å. The size of the simulation box containing pristine TiO2-supported Au nanoparticles is 20 Å × 20 Å × 30 Å, being larger than the realistic size of the composite system.
Modeling TiO2-supported Au nanoparticles
Calculated surface energies (in J/m2) for the anatase (0 0 1) and (1 0 1) surfaces, calculated based on GGA pseudo-potential
(0 0 1)
(1 0 1)
Results and discussion
Structural parameters and adsorption energies
Bond lengths (in Å) and angles (in degrees) of NO2 molecule adsorbed on the TiO2-supported Au nanoparticles
Adsorption energies (in eV) and Mulliken charge values (in e) for NO2 molecule adsorbed on bare TiO2, bare Au, and TiO2-supported Au overlayers
Type of complex
Adsorption energy (eV)
Mulliken charge (normal basis sets)
Mulliken charge (large basis sets)
Therefore, the adsorption of NO2 on the TiO2-supported Au nanoparticle (configuration A) is more favorable in energy than the adsorptions in configurations B and C. For both adsorption types A and C, the adsorption energy is higher (more negative) than the adsorption energy of adsorption type B. The reason is that the configurations A and C provide a double contacting point between the NO2 and TiO2-supported Au, whereas configuration B shows a single contacting point. NO2 molecule was strongly coordinated to the TiO2-supported Au by its two oxygen atoms. In other words, two oxygen atoms of the NO2 molecule can interact with the TiO2-supported Au overlayer more efficiently. The adsorption energies calculated from DFT-D2 and DFT-D3 methods are significantly larger than those obtained from the standard DFT calculations. Tamijani et al.  reported the results of the adsorption of noble-gas atoms on the TiO2 (110) surface-based van der Waals corrected DFT approach and clearly demonstrated the increase in the adsorption energies caused by vdW interactions.
The adsorption energies are considerably increased when the adsorption energies are corrected for dispersion energy, as shown in Table 3. We have calculated the adsorption energies for NO2 molecule on the bare Au and TiO2 nanoparticles. As can be seen from Table 3, the adsorption energy of NO2 molecule on the Au nanoparticle is about −1.23 eV and that of pristine TiO2 is −0.72 eV, and NO2 adsorption on the TiO2-supported Au nanocomposite is found to be −1.48 eV. The higher adsorption energy gives rise to a strong interaction between the adsorbent and adsorbed molecules, and its more negative sign also represents an energy favorable process. Thus, NO2 adsorption on the TiO2-supported Au nanocomposite is more energetically favorable than the adsorption on the bare Au and TiO2 nanoparticles. In the TiO2-supported Au overlayers, the interactions of NO2 and TiO2 are stronger than those between NO2 and bare TiO2 nanoparticles, indicating that Au nanoparticle is conducive to the interaction of NO2 molecule with TiO2 nanoparticles. In other words, Au nanoparticle enhances NO2 detection by means of the TiO2-supported Au nanocomposite-based sensors. Therefore, the results of adsorption energies suggest that the TiO2-supported Au nanoparticle is a good candidate to be utilized in sensing of toxic NO2 molecules in the environment.
Mulliken charge analysis
For instance, configuration A represents a sizeable charge transfer of about 0.15 |e| (e, the electron charge) from the TiO2-supported Au nanoparticle to the NO2 molecule, implying that NO2 plays an important role as a charge acceptor. It is worth mentioning that the charge exchange between adsorbent and adsorbed molecule affects the conductivity of the system, being a great strategy to design more efficient and more appropriate sensor devices for the detection of NO2 in the environment.
One of the most important problems in Mulliken charge analysis is the intense sensitivity of Mulliken charges to the basis set choice. Fundamentally, a comprehensive basis set for a molecule can be covered by placing a large set of functions on a single atom. In the Mulliken scheme, all the electrons would then be assigned to the single atom. Therefore, it is well known that the Mulliken charge approach has no complete basis set limit, as the precise value which strongly depends on the way the limit is approached. Consequently, an efficient convergence for charges does not exist, and different basis set families may produce extremely different results. To overcome this problem, many modern approaches can be tried for estimating net atomic charges, such as electrostatic potential and natural population methods . We have also calculated the Mulliken charges with the large basis sets of higher accuracy and then found that increasing basis set can simply modify Mulliken charges by approximately 0.22 e. The obtained results are summarized in Table 3. This table shows that the strong basis sets give rise to an increase in the Mulliken charge values.
DFT calculations were carried out to investigate the sensing performance of undoped TiO2-supported Au nanoparticles for the detection of NO2 molecules. The adsorption behaviors of NO2 on the TiO2-supported Au nanoparticles were investigated in detail. The results show that the N–O bonds of the adsorbed NO2 molecule were elongated after the adsorption process, which indicates the weakening N–O bonds of the NO2 molecule. The results also suggest that the interaction of the NO2 molecule with the TiO2-supported Au overlayer through its oxygen atoms is energetically more favorable than the interaction of NO2 through its nitrogen atom. This interaction provides a double interacting point between the NO2 and TiO2-supported Au, suggesting the strong adsorption of NO2 over the substrate. The current results suggest that Au nanoparticle in the TiO2-supported Au nanocomposites affects the final configuration of TiO2 nanoparticles with adsorbed NO2 molecules and, therefore, strengthens the interaction between NO2 and TiO2. The substantial overlaps between the PDOSs of the Au and oxygen atoms, as well as, Au and nitrogen atoms indicate the formation of chemical bonds between them. Mulliken population analysis reveals a noticeable charge transfer from the TiO2-supported Au to the NO2 molecule, indicating the acceptor characteristic of the NO2 molecule. Based on the inclusion of vdW interactions, we found that the adsorption energies become larger. However, our findings thus suggest that the TiO2-supported Au nanoparticles can be utilized as potentially efficient gas sensors for NO2 recognition.
This work was supported by Azarbaijan Shahid Madani University.
- 2.Fujishima, A., Hashimoto, K., Watanabe, T.: TiO2 photocatalysis: fundamentals and applications. Bkc, Tokyo (1999)Google Scholar
- 4.Abbasi, A., Sardroodi, J.J.: Modified N-doped TiO2 anatase nanoparticle as an ideal O3 gas sensor: insights from density functional theory calculations. J. Comp. Theor. Chem. 600, 2457–2469 (2016)Google Scholar
- 6.Batzilla, M., Morales, E.H., Diebold, U.: Surface studies of nitrogen implanted TiO2. J. Chem. Phys. 339, 36–43 (2007)Google Scholar
- 36.Ozaki, T., Kino, H., Yu, J., Han, M.J., Kobayashi, N., Ohfuti, M., Ishii, F., et al.: The code OpenMX, pseudoatomic basis functions, and pseudopotentials are available on a web site ‘http://www.openmxsquare.org’ (2017). Accessed 2 Mar 2017
- 44.Downs, R.T.: Web page at: http://rruff.geo.arizona.edu/AMS/amcsd.php (2014). Accessed 9 May 2014
- 45.Wyckoff, R.W.G.: Crystal structures, 2nd edn. Interscience Publishers, New York (1963)Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.