This is the submitted version of the journal article: Epifani, Mauro; Kaciulis, Saulius; Mezzi, Alessio; [et al.]. «Rhodium as efficient additive for boosting acetone sensing by TiO2 nanocrystals. Beyond the classical view of noble metal additives». Sensors and actuators. B, Chemical, Vol. 319 (Sep. 2020), art. 128338. DOI 10.1016/j.snb.2020.128338 This version is available at https://ddd.uab.cat/record/236005 under the terms of the license *Revised Manuscript Click here to view linked References Rhodium as Efficient Additive for Boosting Acetone Sensing by TiO2 nanocrystals. Beyond the Classical View of Noble Metal Additives Mauro Epifani[1],a, Saulius Kaciulis[2], Alessio Mezzi[2], Ting Zhang,[3] Jordi Arbiol[3,4], Pietro Siciliano[1], Anton Landström[5], Isabella Concina[5], Abderrahim Moumen[6], Elisabetta Comini[6], Chu Xiangfeng[7] 1 Istituto per la Microelettronica e i Microsistemi, IMM-CNR, Via Monteroni, 73100 Lecce, Italy; amaurosalvatore.epifani@cnr.it 2 Istituto per lo Studio dei Materiali Nanostrutturati, ISMN–CNR, PO Box 10, 00015 Monterotondo Stazione, Roma, Italy; 3 Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC BIST, Campus UAB, Bellaterra, 08193 Barcelona, Catalonia, Spain 4 ICREA, Pg. Lluís Companys 23, 08010 Barcelona, Catalonia, Spain 5 Department of Engineering Science and Mathematics, Luleå University of Technology, 6 Department of Information Engineering, Brescia University, Via Valotti 9, 25133 Brescia, Italy 7 School of Chemistry and Chemical Engineering, Anhui University of Technology, Maanshan 243002, Anhui, P. R. China 1 Abstract Anatase TiO2 nanocrystals were prepared by solvothermal synthesis and modified by in- situ generated Rh nanoparticles, with a starting nominal Rh:Ti atomic concentration of 0.01 and 0.05. After heat-treatment at 400 °C the TiO2 host was still in the anatase crystallographic phase, embedding Rh nanoparticles homogeneously distributed and whose surface had been oxidized to Rh2O3, as established by X-ray diffraction, Transmission Electron Microscopy and X-ray Photoelectron spectroscopy. Moreover, Rh seemed also homogeneously distributed in elemental form or as Rh2O3 nanoclusters. The acetone sensing properties of the resulting materials were enhanced by Rh addition, featuring a response increase of one order of magnitude at the best operating temperature of 300 °C. Moreover, Rh addition enlarged the detection range down to 10 ppm whereas pure TiO2 was not able of giving an appreciable response already at a concentration as high as 50 ppm. From the sensing data, the enhancement of the sensor response was attributed to the finely dispersed Rh species and not to the oxidized Rh nanocrystals. Keywords: TiO2 nanocrystals, acetone sensing, rhodium, noble metal additives, solvothermal synthesis. 2 Introduction Already for a long time noble metal additives have been used as performance promoters of chemoresistive, metal oxide based gas sensors [1-3]. Such effect is classically [4] attributed to spillover of adsorbed species onto the oxide semiconductor support or Fermi level control. Indeed, loading with noble metals has been exploited already in the early commercial gas sensors fabricated by Figaro Engineering Inc. , where Pd was traditionally added to the oxide host. Along with Pd, even Pt has been often employed as additive [5, 6]. Much less attention has been paid to another noble metal group element such as rhodium. Only few examples are known of Rh addition to Ga2O3 [7], SnO2 [8, 9], In2O3 [10], WO3 [11]. Interestingly, it has been recently pointed out [12, 13] that the effectiveness of Rh addition is due to Fermi level control upon contact of SnO2, In2O3 and WO3 host with Rh nanoparticles oxidized to Rh2O3. To date, there are no studies on the effect of Rh addition to TiO2 with the exception of a single paper about the response to oxygen [14]. This combination is, however, interesting since the sensing performance of anatase TiO2 is known to benefit from noble metal loading [15-19]. TiO2 is, therefore, a suitable host for investigating and understanding how sensing is positively affected by the Rh additive. This paper demonstrates the beneficial effect of Rh addition to the sensing properties of TiO2 nanocrystals. The Rh additive was present in several different structures, large rhodium nanoparticles in which only the surface is oxidized and small homogeneously dispersed Rh species. On this basis, the actual importance of electronic sensitization by Rh2O3 was discussed in comparison with the possible effect of other Rh species present on the TiO2 host, such as elemental Rh and/or Rh2O3 nanoclusters. application-aimed but is serves to show how the effects of noble metal additives should be investigated when dealing with such complex systems. 3 Experimental Pure TiO2 nanocrystals were prepared by first precipitating 4 mL of TiO2 sol in 10 mL of n-dodecylamine (DA, 98%, Sigma-Aldrich), followed by heating for 1 h at 100 °C in a glass vial. The TiO2 sol was prepared by solvolysis of 1.5 g TiCl4 in 10 mL of methanol, followed, after cooling, by the addition of 2.4 mL of water. The precipitate obtained in DA was extracted by methanol and washed two times in acetone, then redispersed in 12 mL of oleic acid (90%, Sigma-Aldrich). The resulting suspension was transferred to a borosilicate glass vial and treated for 2h at 250 °C into a steel autoclave (45 mL). After cooling, and without any purification, the prescribed amount of Rh(acac)3, dissolved in 1 mL of chloroform, was added, followed by another heating step in the same autoclave at 250 °C for 1h. Two nominal Rh concentrations were considered, equal to 1% and 5% atomic concentration with respect to Ti. In the following, such samples will be indicated as TiO2- 1Rh and TiO2-5Rh, respectively. The final product was extracted by methanol, washed 3 times with acetone and then dried at 90 °C in air, after which a black powder was obtained. Heat-treatment of the samples at 400 °C for 1 h was carried out in a muffle furnace at a rate of 5 °C/min, after placing them into a glazed porcelain crucible. After the heat-treatment, a black powder was again obtained. X-ray diffraction (XRD) patterns were recorded on a PANalytical Empyrean diffractometer. Cu K-α radiation was obtained with a Cu anode operated at 45 kV/40 mA and a graphite monochromator. A programmable divergence slit was used to illuminate a 10x10 mm2 area of the sample surface. Rietveld refinement of the patterns was carried out by the MAUD software [20]. The X-ray photoelectron spectroscopy (XPS) analyses were carried in an Escalab MkII (VG Scientific, East Grinstead, UK) electronic spectrometer equipped with unmonochromatized Al/Mg source and a 5-channeltrons detector. The samples were pressed on the grated Au 4 foil (99.99%) fixed on the standard Escalab holder stubs. Large area XPS measurements were performed at 20 eV pass energy. The binding energy (BE) scale was calibrated by measuring the reference peak of Au4f7/2 (84.0 ± 0.1 eV) from the supporting foil. Experimental data were processed by using the software Avantage v.5 (Thermo Fisher Scientific). High-resolution TEM (HRTEM) and scanning TEM (STEM) studies were carried out using a field emission gun FEI Tecnai F20 microscope at 200 kV with a point-to-point resolution of 0.19 nm. High angle annular dark-field (HAADF) STEM was combined with electron energy loss spectroscopy (EELS) in the Tecnai microscope by using a GATAN QUANTUM filter. The sensing devices were processed onto alumina substrates (2x2 mm2) provided with Ti/Pt electrodes and heaters. A paste was prepared by mixing the as-prepared samples with 1,2- propanediol. Heat-treatment of the paste at 400 °C was carried out in the device itself, by using the integrated heaters. The gas-sensing tests were carried out in flow-through mode with dry air as gas carrier [21]. The desired gas concentration was obtained by mixing with mass flow controllers the target gas with dry air. The response of the devices was defined as |Ggas-G0|/G0, where G0 was the sensor baseline electrical conductance in dry synthetic air and Ggas indicated the sensor electrical conductance after dynamic equilibration with the target gas. The sensor signal readout was carried out with a picoammeter (Keithley, 6485). A fixed voltage (1V) was applied to the sensors. Results and Discussion A. General considerations on the synthesis procedure. Structure and composition of the samples Two main issues were considered in the synthesis design, the choice of the Rh precursor and the addition procedure to the TiO2 host material. Previous work [22] on noble metal 5 precursors had shown that metal acetylacetonates are a preferred choice with respect to ordinary salts, due to more controlled decomposition and nanoparticle growth and dispersion into the oxide host. Rhodium was no exception, since RhCl3 effectively resulted into the formation of Rh nanoparticles but also in vigorous reaction into the autoclave, with the suspension splashed all over the autoclave walls. For reacting Rh(acac)3 with the TiO2 suspension, its addition was also attempted without any preliminary solvothermal treatment of pure TiO2. Nevertheless, as shown in the Supporting Information (Figure S1), a few unidentified peaks appeared in the related XRD pattern, which did not occur with the finally chosen procedure described in the Experimental. After clarifying these issues, the crystal structure and chemical composition of the materials was investigated. Figure 1 shows the comparison between the XRD patterns of the as-prepared and heat-treated samples. The reflections of anatase TiO2 are clearly present in the patterns, with obvious narrowing after heat-treatment due to grain growth. It can be seen that the signal of elemental Rh only appears for the TiO2-5Rh sample, while for the TiO2-1Rh sample the effective Rh presence was ensured by XPS (see Supporting Information, Figure S2). Figure 1: XRD patterns of the indicated samples. Only the (111) Rh reflection is marked, all the other reflection belonging to anatase TiO2. 6 There is no evidence of Rh2O3 reflections and the Rh peak intensity after heat-treatment at 400 °C decreased with respect to the as-prepared sample. In the Supporting Information the Rietveld refinement results are shown of both the as-prepared and heat-treated TiO2-5Rh samples (Figures S3 and S4). The TiO2 lattice parameters were in full agreement with the values reported for anatase. Moreover, the TiO2 crystal domain size increased from 6.5 to 11.4 nm after the heat-treatment at 400 °C, while the Rh grain size increased from 2.4 to about 8.2 nm. Despite the low Rh concentration made the quantitative determination highly uncertain, further refinement by the diffractometer proprietary software confirmed the obtained trends. The Rietveld analysis also provided the Rh weight %, which decreased from 0.683 to 0.155% after the heat-treatment, in agreement with the qualitative evaluation of the XRD patterns. The actual atomic concentration of nanocrystalline Rh with respect to Ti was therefore calculated as 0.12 %. Also shown in the Supporting Information is the refinement of the XRD patterns of pure TiO2 heat-treated at 400 °C. The obtained TiO2 size of 7.8 nm was even lower than in the TiO2-5Rh sample, which could also be due to artifacts of the Rietveld refinement or also indicative of Rh catalytic effect on the removal/decomposition of oleic acid ligands. Anyway even simple calculation of the grain size by Scherrer’s equation provided similar results, despite they were more comparable (6.3 nm for pure TiO2, 7.7 nm for TiO2-5Rh). What was important from these size determinations was that any sensing difference between pure and Rh-TiO2 could not be ascribed to any size effect. XPS spectra recorded for the 400 °C TiO2-5Rh sample are shown in Figure 2. The related quantification results are reported in Table 1. 7 Table 1: Elemental quantification of the TiO2-5Rh sample after heat-treatment at 400 °C. Peak BE, eV FWHM, eV Atomic % Chemical State C1s A 285.0 2.1 11.1 aliphatic C1s B 286.8 2.1 2.8 C-O C1s C 289.3 2.1 1.0 carboxyl O1s A 530.4 1.9 50.9 oxide O1s B 532.2 1.9 8.1 OH- Rh3d 308.3 2.2 0.2 Rh2O3 Ti2p 459.1 1.9 25.9 TiO2 Figure 2: XPS spectra of the main peaks for the TiO2-5Rh sample heat-treated at 400 °C. 8 No chlorine (possible residual of TiCl4) was detected but a negligible concentration of carboxylic carbon (possible fingerprint of oleic acid residuals) was present. The O 1s signal was composed only of the expected hydroxyl and oxide components. The 0.77 % Rh:Ti concentration ratio was lower than the nominal 5% value, consequently to the loss of Rh during the purification step. The supernatant recovered after centrifugation was indeed deep brown colored, indicating that a part of Rh had not been retained in the TiO2 matrix. Another sample was prepared and analyzed, and the same Rh concentration was detected. Instead, XPS detection of Rh in the TiO2-1Rh samples was very hard (see Figure S2), so in the rest of the work any consideration will be referred to the TiO2-5Rh samples and devices. It is interesting that only Rh(III) was consistently detected in all of the analyzed samples, indicating bonding with oxygen as in Rh2O3. No Rh nanocrystals could be singled out in the TEM images, many of which were analyzed, as shown in the Supporting Information (Figure S5). As an example, Figure 3 shows a collection of HRTEM images from the 400 °C TiO2-5Rh sample. The treatment removed the oleic acid capping, resulting in the observed aggregation of nanocrystals. Nevertheless, it was still possible to confirm a typical size around 10 nm, in agreement with the previously reported result of the Rietveld refinement. The inset shows a magnification of the remarked region and the related power spectrum. The analyzed nanoparticles, like many others investigated in the sample, displayed lattice parameters typical of anatase TiO2, in full agreement with the XRD data. Once again, there was no evidence of Rh incorporation into the TiO2 lattice. 9 Figure 3: HRTEM micrograph of the 400 °C TiO2-5Rh sample, detail of the orange squared region and its corresponding power spectrum. More interesting is the observation of the EELS mapping shown in Figure 4. It is evident that the Rh distribution follows quite closely that of Ti and of O. Summarizing: i) the XRD results in Figure 1 only showed the reflections of Rh nanocrystals. ii) In XPS spectra only Rh (III) was detected. The XPS Rh : Ti atomic concentration was 0.77 %, larger than that obtained from XRD (0.12%). Therefore Rh in form of nanocrystals was only a small fraction of the overall Rh content. iii) EELS showed homogeneous Rh distribution but it was not possible to detect Rh nanocrystals in many analyzed images. 10 Figure 4: EELS chemical composition maps obtained from the red squared area of the STEM micrograph for the 400 °C TiO2-5Rh sample. Individual Ti L2,3-edges at 456 eV (red), O K-edge at 532 eV (blue) and Rh M4,5-edges at 307 eV (green) as well as the composite (Ti-O, Rh-O and Ti-Rh) elemental maps are shown. For putting together all of these results, it was concluded that in the 400 °C TiO2-5Rh sample there were two Rh species: i) a prevalent distribution of oxidized Rh species dispersed homogeneously, most probably small species comprising a few Rh atoms, not detectable by TEM; in the following these species will be referred to as “Rh oxide nanoclusters” despite this nomenclature may be not completely appropriate. For instance it is not excluded that single Rh cations may be incorporated into the surface lattice of anatase. ii) a few Rh nanocrystals, with a size comparable to the TiO2 host, and whose surface was oxidized to Rh(III) to such an extent that the XPS probe did not get to the Rh core, which was instead detected by XRD. In the following these species will be referred to as “surface oxidized nanocrystals”. The resulting compositional and structural status of the samples made very complex to ascertain how Rh may influence the gas-sensing properties. Fortunately, as discussed in the 11 following section, comparison with recent literature results allowed proposing some working hypotheses. B. Gas-sensing properties Figure 5 shows the comparison between the dynamic response curves to acetone of pure TiO2 and the 400 °C TiO2-5Rh sample at an operating temperature of 300 °C (the data for other temperatures are shown in the Supporting Information, Figure S6). The response improvement when Rh was added to TiO2 is obvious. The same holds for ethanol response, as shown in the Supporting Information (Figure S8). It can also be noted that TiO2 poorly reacts to acetone concentration changes. Figure 5: Dynamic response curves to acetone at 300 °C of TiO2 and TiO2-5Rh devices. The calibration data in Figure 6 show that the acetone response is enhanced by almost one order of magnitude for all the tested concentrations. The best operating temperature was 300 °C, as demonstrated in Figure 7. The response data for the various acetone concentrations followed the typical volcano curve. The sensing data as such clearly demonstrated the beneficial effect of Rh addition, and in this sense the results discussed until now are in line with the well-established effect of noble metals on gas-sensing properties. On the other hand, it has been recently pointed out [12] that Rh (oxidized to 12 Rh2O3) may also have a detrimental effect on acetone and NO2 sensing properties of In2O3, SnO2 and WO3, while the response of WO3 to CO and ethanol was improved. Figure 6: acetone calibration curves of TiO2 and TiO2-5Rh devices at an operating temperature of 300 °C. In the present case, response improvement was observed (also for ethanol, see Supporting Information, Figure S9). It was concluded that it must be due to a combined effect by the various Rh species. In which way would such species interact with the support to provide response improvement? Figure 7: comparison between the TiO2-5Rh acetone responses at various operating temperatures. The typical volcano behavior can be observed. For investigating this point, conductance measurements were carried out in increasing concentrations of oxygen in nitrogen. The results are shown in Figure 8. There are some 13 striking differences between the two series of samples. First of all the TiO2-5Rh sensor showed much lower conductance than pure TiO2 in analogous conditions, suggesting that Rh species resulted in charge extraction from TiO2. More importantly, the TiO2-5Rh series is much less sensitive to changes in the oxygen concentration, above all at temperatures from 200 to 400 °C, when oxygen ionosorption becomes more and more important, resulting in conductance decrease. This result suggested that electronic control was carried out by Rh species onto the sensing properties, so detection of acetone (and ethanol and other analytes, see the Supporting Information) is favored by charge restoring from such species to the TiO2 lattice upon interaction with the gases. This conclusion would seem contradictory with ref. 12 since acetone response in the present case has been enhanced by Rh addition. But: in the present work, the Rh surface oxidized nanocrystals are not present in high concentration, as explained above. Figure 8: Conductance values of the indicated devices measured as a function of the oxygen concentration and operating temperature in dry nitrogen. Second, since only their surface is oxidized, quite a different situation from ref. 10 must be considered. Third, the size of Rh surface oxidized nanocrystals and TiO2 nanocrystals is comparable, which is a different situation from the classical view of Fermi level pinning. Even more importantly, Rh is largely present as oxide nanoclusters, as evidenced above in 14 the discussion of the EELS data. Moreover, the ex situ characterization of the materials did not allow establishing the actual oxidation state of Rh in operando conditions. Finally, it was excluded that extensive doping of TiO2 had occurred, as elaborated in detail in the Supporting Information. For putting together all these facts, it was concluded that the sensor response is still dominated by exchange of charges with the TiO2 support, also occurring by Rh oxide nanoclusters. For such species, the exchange of charges would be the analogue of Fermi level control, despite the Fermi level cannot be defined for such non-bulk, non-crystalline species [6]. Why even acetone response is enhanced, not only that to ethanol, is not clearly attributable to a specific Rh additive species. In fact, analyzing the various contributions to the response does not appear as straightforwardly feasible. In ref. 6 the authors defined the presence of PtOx nanoclusters, which could also be evidenced by TEM as flat structures attached to the surface of SnO2 support. In the present work, the hypothesized presence of few-atoms RhOx species, still defined as nanoclusters, ultimately raises the level of difficulty when analyzing the noble metal effect onto the sensing properties of a given oxide support. The last sentence is meant as the main indication to future workers in the field: are tiny, dispersed species (“nanoclusters”) present when using noble metal additives in the processing of gas sensors? In this case, how to deal with such species? While it is tempting to treat their effect analogously to the classical promotion mechanisms, their size and structure fundamentally hinders such approach. As we have seen above, the difficulty in defining a Fermi level for them hinders the electronic sensitization mechanism to be immediately applied. But also in the case of spillover it would be difficult to suppose a steady reservoir of adsorbed analyte species without any concurrent interaction with sensing semiconductor. Conclusions 15 It is possible to homogeneously add Rh to anatase TiO2 nanocrystals by suitably controlling the Rh precursor and the overall processing route. The heat treatment temperature can determine a complex distribution of Rh species, where aside surface oxidized Rh nanocrystals a distribution of Rh oxide nanoclusters exists. From the analysis of the sensing tests and comparison with literature reports, it has emerged that this distribution of Rh oxide nanoclusters can be responsible for remarkable enhancement of the sensor response with respect to pure TiO2. It is necessary then, when establishing the effect of noble metal additives, to determine in fine detail the effective speciation of the noble metal. As a consequence, the classical concepts of electronic sensitization and spillover must be cautiously used and possibly rethought. Acknowledgements ICN2 acknowledge funding from Generalitat de Catalunya 2017 SGR 327 and the Spanish MINECO project ENE2017-85087-C3. ICN2 is supported by the Severo Ochoa program from Spanish MINECO (Grant No. SEV-2017-0706) and is funded by the CERCA Programme / Generalitat de Catalunya. Part of the present work has been performed in the framework of Universitat Autònoma de Barcelona Materials Science PhD program. TZ has received funding from the CSC-UAB PhD scholarship program. Competing interests statement The authors have no competing interests to declare. References [1] N. Yamazoe, New Approaches for Improving Semiconductor Gas Sensors, Sens. Actuators, B 5 (1991) 7-19. 16 [2] D. Degler, U. Weimar, N. Barsan, Current Understanding of the Fundamental Mechanisms of Doped and Loaded Semiconducting Metal-Oxide-Based Gas Sensing Materials, ACS Sensors 4 (2019) 2228-2249. [3] S.A. Müller, D. Degler, C. Feldmann, M. Türk, R. Moos, K. Fink, F. Studt, D. Gerthsen, N. Bârsan, J.-D. Grunwaldt, Exploiting Synergies in Catalysis and Gas Sensing using Noble Metal-Loaded Oxide Composites, ChemCatChem 10 (2018) 864-880. [4] N. Yamazoe, Y. Kurokawa, T. Seiyama, Effects of additives on semiconductor gas sensors, Sens. Actuators, B 4 (1983) 283-289. [5] Y.S. Liu, X. Gao, F. Li, G.Y. Lu, T. Zhang, N. Barsan, Pt-In2O3 mesoporous nanofibers with enhanced gas sensing performance towards ppb-level NO2 at room temperature, Sens. Actuators, B 260 (2018) 927-936. [6] D. Degler, S.A. Muller, D.E. Doronkin, D. Wang, J.D. Grunwaldt, U. Weimar, N. Barsan, Platinum loaded tin dioxide: a model system for unravelling the interplay between heterogeneous catalysis and gas sensing, J. Mater. Chem. A 6 (2018) 2034-2046. [7] A.C. Lang, M. Fleischer, H. Meixner, Surface modifications of Ga2O3 thin film sensors with Rh, Ru and Ir clusters, Sens. Actuators, B 66 (2000) 80-84. [8] G. Korotcenkov, V. Brinzari, P. Hanys, V. Nehasil, XPS study of the SnO2 films modified with Rh, Surf. Interface Anal. 50 (2018) 795-801. [9] M. Anton, B. Budy, The Influence of Rhodium on SnO2-CO Gas Sensors, Sens. Actuators, B 19 (1994) 500-501. [10] S.J. Kim, I.S. Hwang, C.W. Na, I.D. Kim, Y.C. Kang, J.H. Lee, Ultrasensitive and selective C2H5OH sensors using Rh-loaded In2O3 hollow spheres, J. Mater. Chem. 21 (2011) 18560-18567. 17 [11] N.H. Kim, S.J. Choi, S.J. Kim, H.J. Cho, J.S. Jang, W.T. Koo, M. Kim, I.D. Kim, Highly sensitive and selective acetone sensing performance of WO3 nanofibers functionalized by Rh2O3 nanoparticles, Sens. Actuators, B 224 (2016) 185-192. [12] A. Staerz, I. Boehme, D. Degler, M. Bahri, D. Doronkin, A. Zimina, H. Brinkmann, S. Herrmann, B. Junker, O. Ersen, J.-D. Grunwaldt, U. Weimar, N. Barsan, Rhodium Oxide Surface-Loaded Gas Sensors, Nanomaterials 8 (2018) 892. [13] A. Staerz, T.H. Kim, J.H. Lee, U. Weimar, N. Barsan, Nanolevel Control of Gas Sensing Characteristics via p-n Heterojunction between Rh2O3 Clusters and WO3 Crystallites, J. Phys. Chem. C 121 (2017) 24701-24706. [14] L. Castaneda, A. Maldonado, M.D. Olvera, Sensing properties of chemically sprayed TiO2 thin films using Ni, Ir, and Rh as catalysts, Sens. Actuators, B 133 (2008) 687-693. [15] J. Bai, B. Zhou, Titanium Dioxide Nanomaterials for Sensor Applications, Chem. Rev. 114 (2014) 10131-10176. [16] H. Liu, W. Shen, X. Chen, A room temperature operated ammonia gas sensor based on Ag-decorated TiO2 quantum dot clusters, RSC Adv. 9 (2019) 24519-24526. [17] A.M. Ruiz, A. Cornet, K. Shimanoe, J.R. Morante, N. Yamazoe, Effects of various metal additives on the gas sensing performances of TiO2 nanocrystals obtained from hydrothermal treatments, Sens. Actuators, B 108 (2005) 34-40. [18] D.M. Tobaldi, S.G. Leonardi, K. Movlaee, L. Lajaunie, M.P. Seabra, R. Arenal, G. Neri, J.A. Labrincha, Hybrid Noble-Metals/Metal-Oxide Bifunctional Nano-Heterostructure Displaying Outperforming Gas-Sensing and Photochromic Performances, ACS Omega 3 (2018) 9846-9859. [19] M. Epifani, T. Andreu, R. Zamani, J. Arbiol, E. Comini, P. Siciliano, G. Faglia, J.R. Morante, Pt doping triggers growth of TiO2 nanorods: nanocomposite synthesis and gas- sensing properties, CrystEngComm 14 (2012) 3882-3887. 18 [20] L. Lutterotti, Total pattern fitting for the combined size-strain-stress-texture determination in thin film diffraction, Nucl. Instrum. Meth. B 268 (2010) 334-340. [21] M. Epifani, S. Kaciulis, A. Mezzi, D. Altamura, C. Giannini, P.Y. Tang, J.R. Morante, J. Arbiol, P. Siciliano, E. Comini, I. Concina, Solvothermal Synthesis, Gas-Sensing Properties, and Solar Cell-Aided Investigation of TiO2-MoOx Nanocrystals, ChemNanoMat 3 (2017) 798-807. [22] M. Epifani, J. Arbiol, E. Pellicer, E. Comini, P. Siciliano, G. Faglia, J.R. Morante, Synthesis and gas-sensing properties of Pd-doped SnO2 nanocrystals. A case study of a general methodology for doping metal oxide nanocrystals, Cryst. Growth Des. 8 (2008) 1774-1778. 19 Supplementary Material Click here to download Supplementary Material: Supporting Information.pdf *Author Biographies Mauro Epifani obtained the PhD degree in Physics in 2000 from the University of Lecce, with a thesis on the chemical preparation of metal and metal oxide nanostructures and their optical and electrical characterization. Since then, his research interests include the preparation, processing and characterization of metal and semiconductor colloids, metal oxide thin films and powders, hybrid organic-inorganic materials for gas sensing, water remediation and energy. Since the end of 2001 he is a researcher at the Italian National Council of Researches (CNR). Saulius Kaciulis received his PhD in 1983 at the University of Vilnius in Lithuania. At present time, he is a senior researcher at the Institute for the Study of Nanostructured Materials of the CNR of Italy. His primary research activities are focused on the surface analysis of innovative materials, such as thin films for thermionic converters and gas sensing, 2D carbon materials, anticorrosion coatings, biocompatible materials, etc. Co-author of over 200 scientific papers. Alessio Mezzi graduated in Chemistry at “La Sapienza” University of Rome, Italy. He received his Ph.D. in Materials Engineering from University of Rome “Tor Vergata”, Italy, in 2004. He is a researcher at the Institute for the Study of the Nanostructured Materials of the CNR, Italy. His main research interest is focused on the study of innovative materials by using the surface analysis techniques (XPS and AES). Ting Zhang received her B. S. in chemical engineering and technology from Shanghai University, China in 2013. And She obtained her master degree in physical chemistry from Lanzhou University, China in 2017. Now, she is a Ph. D candidate in Prof. Jordi Arbiol’s group at the Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC & BIST. Her research is focused on preparation and TEM characterization of catalytic materials for electrocatalysis. Prof. Jordi Arbiol graduated in Physics at Universitat de Barcelona (UB) in 1997, where he also obtained his PhD (European Doctorate and PhD Extraordinary Award) in 2001 in the field of transmission electron microscopy (TEM) applied to nanostructured materials. He was Assistant Professor at UB. From 2009 to 2015 he was Group Leader at Institut de Ciència de Materials de Barcelona, ICMAB-CSIC and Scientific Supervisor of its electron microscopy facilities. Since 2017 he is President of the Spanish Microscopy Society (SME), was Vice-President from 2013 to 2017 and since 2009 he is Member of its Executive Board. In 2018 he was elected as Member of the Executive Board of the International Federation of Societies for Microscopy (IFSM) (2019-2026). Since 2015 he is the leader of the Group of Advanced Electron Nanoscopy at Institut Català de Nanociència i Nanotecnologia (ICN2), CSIC and The Barcelona Institute of Science and Technology (BIST). Hes has published more than 330 peer-review journals with more than 12750 citations and h-index 64 WoS (74 GoS). Pietro Siciliano, physicist, senior researcher, received his degree in physics in 1985 from the University of Lecce. He took his PhD in physics in 1989 at the University of Bari. He is currently a senior member of the National Council of Research in Lecce, where he has been working for many years in the field of preparation and characterization of thin film for gas sensor, being in charge of the sensors group. Currently he is director of IMM-CNR in the Department of Lecce. Anton Landström received his M. Sc. degree in 2017 from Luleå University of Technology, where he is currently working as a doctoral student in experimental physics at the department of engineering sciences and mathematics. His research is focused on the synthesis and characterization of advanced nanostructured materials for use in energy harvesting applications. Isabella Concina earned her PhD in Chemical Sciences at the University of Padova (Italy) in 2006 and she currently is associate professor in experimental physics at Luleå University of Technology, Sweden. She is a Marie Curie Fellow under the VINNOVA program Mobility for Growth. She has been previously working as assistant professor at the University of Brescia (Italy) and as research fellow within the National Research Council (Italy). Her main research interests are focussed on the design of nanosized semiconducting metal oxides for structure-driven functional applications, especially dealing with light-matter interaction, such as solar energy conversion and photocatalysis. Abderrahim Moumen got his master’s degree in physics and chemistry of materials and nanomaterials at the University of Casablanca, Department of physics. Currently, he is a PhD student in Sensor Laboratory, Brescia University-Italy since November 2017, and his research activities are mainly focused on the preparation and characterization of Metal Oxides (MOXs) nanostructures for chemical sensors applications. The global aim of his research is elaborating or structuring materials at the nanoscale which will open new possible applications such as chemical sensors and gas sensors in particular. Elisabetta Comini received her degree in physics at the University of Pisa in 1996. She received her Ph.D. degree in material science at the University of Brescia. In 2001 she has been appointed assistant professor of physics of matter at Brescia University. In 2016 she became full professor. She has been organizer of several symposia in the sensing field for MRS and E-MRS. She was chair of MRS fall meeting 2013. She has a high productivity confirmed by the numerous publications on international journals (more than 350) and the high number of invited presentations at conferences. EC is a researcher specialist in the growth of metal oxides, particularly nanowires, thin films and the measurement of their electronic, functional and structural properties. EC is the director of SENSOR laboratory (Brescia University, http://sensor.unibs.it). Her Hirsch index (h-index) is 51 (Web of Science), 53 (Scopus), 61 (Google scholar). ORCID http://orcid.org/0000-0003-2559-5197 Google scholar: https://scholar.google.it/citations?user=2ZAnRXwAAAAJ&hl=en Xiangfeng Chu is a professor in School of Chemistry and Chemical Engineering, Anhui University of Technology. He obtained Ph. D degree from University of Science and Technology of China in 1999, and did postdoctoral work in Lecce University (1999.10-2000.09) and Ritsumeikan University (2000.10-2002.04). He ever worked in Sun Yat-sen University (associate professor, 2002.06-2007.01), North Dakota State University (visiting scholar, 2006.04-2006.09) and Clarkson University (visiting scholar, 2006.10-2007.10). His research fields are chemical sensors, chemical mechanical planarization and utilization of fly ash. *Credit Author Statement Conceptualization M. Epifani M. Epifani, S. Kaciulis, A. Mezzi, T. Zhang, J.Arbiol, A. Landström, I. Investigation Concina, A. Moumen, E. Comini, C. Xiangfeng M. Epifani, S. Kaciulis, A. Mezzi, T. Zhang, J.Arbiol, P. Siciliano, A. Visualization Landström, I. Concina, A. Moumen, E. Comini, C. Xiangfeng Writing - Original M. Epifani Draft Writing - Review & M. Epifani, S. Kaciulis, A. Mezzi, T. Zhang, J.Arbiol, P. Siciliano, A. Editing Landström, I. Concina, A. Moumen, E. Comini, C. Xiangfeng