Low-Temperature Catalytic CO Oxidation Over Non-Noble, Efficient Chromia in Reduced Graphene Oxide and Graphene Oxide Nanocomposites

Herein, bare chromia nanoparticles (Cr2O3 NPs) and chromia supported on reduced graphene oxide (rGO) and graphene oxide (GO) hybrids were synthesized, followed by characterization by means of FESEM, Raman spectroscopy, TGA, XRD, TEM/HRTEM, XPS and N2 sorptiometry. The investigated bare Cr2O3 and the hybrids (Cr2O3/rGO and Cr2O3/GO) were employed as catalysts for low-temperature CO oxidation. Compared with the other catalysts, the results revealed efficient catalytic activity using Cr2O3/GO, which was attributed to its higher surface area together with the mixed oxidation state of chromium (Cr3+ and Cr>3+). These are important oxidation sites that facilitate the electron mobility essential for CO oxidation. Moreover, the presence of carbon vacancy defects and functional groups facilitate the stabilizing of Cr2O3 NPs on its surface, forming a thermally stable hybrid material, which assists the CO oxidation process. The Cr2O3/GO hybrid is a promising low-cost and efficient catalyst for CO oxidation at low temperatures. The higher activity of graphene oxide supported Cr2O3 NPs can provide an efficient and cost-effective solution to a prominent environmental problem.


Introduction
Carbon monoxide (CO) is a toxic and highly poisonous gas and is considered a silent killer that contributes indirectly to different environmental and health problems [1]. The main sources of CO are gases emitted from petroleum, chemical and petrochemical industries, and automobile emissions [1]. To overcome its toxicity, CO catalytic oxidation into less harmful carbon dioxide (CO 2 ) is an important strategy. Many researchers have employed different noble metals such as Au, Pt and Pd as potential catalysts for CO oxidation; however, the issues of cost, poor stability, and sensitivity to impurities for deactivation limit their industrial applications [2]. Recently, low-temperature CO oxidation on non-noble metal catalysts has been an important research objective in order to overcome high cost and realize efficient results [3][4][5].
Metal oxide catalysts are regarded as promising materials for industrial and environmental CO oxidation reactions [6]. Their abundance, low cost and pronounced chemical stability make them a promising alternative to noble-metal based catalysts [7][8][9]. Studies that have hitherto been carried out Figure 1A,B shows the FESEM micrographs obtained for GO and rGO. It is clear that, after oxidation, the GO and rGO sheets were both crumpled. The reduction of the graphite sheets due to the distortion of their conjugated π-bond system is also clear. However, a flaky layered microstructure was only shown for GO ( Figure 1A), whereas in rGO, the surface morphology was relatively close ( Figure 1B) due to GO exfoliation during ultrasonic treatment. Raman spectra obtained for GO and rGO are compared in Figure 1C. The two broad peaks therein monitored are the D and G bands, which are the predominant features for carbon-based materials [25]. In the spectrum of GO, the peak at 1589 cm −1 is due to the G band, whereas the peak at 1352 cm −1 is attributed to the D band [25]. The spectrum of rGO shows the G peak at 1582 cm −1 and the D peak at 1348 cm −1 . These peak red-shifts are considered indicative of the recovery of sp 2 domains [26] and, hence, the reduction of GO to rGO. TG curve obtained for the as-obtained chromia, i.e., the un-calcined Cr2O3.4H2O, displays two main decomposition steps, as shown in Figure 2. The first step is shown to undertake about a 25% mass loss maximized at 237 °C. This process is consistent with the removal of weakly held water  TG curve obtained for the as-obtained chromia, i.e., the un-calcined Cr 2 O 3 .4H 2 O, displays two main decomposition steps, as shown in Figure 2. The first step is shown to undertake about a 25% mass loss maximized at 237 • C. This process is consistent with the removal of weakly held water molecules plus the dehydration to Cr 2 O 3 .H 2 O (=2CrOOH). The second step is shown to maximize at 415 • C, giving rise to mass loss of about 11%, which is very close to the dehydration of Cr 2 O 3 .H 2 O into Cr 2 O 3 (10.6%). These results verify the chemical composition of the parent chromium hydroxide compound and emphasize the appropriateness of the calcination temperature regime (≥400 • C), applied for the thermal genesis of the Cr 2 O 3 NPs from the parent hydroxide. Accordingly, we chose the calcination temperature of 600 • C.

Characterization of As-Prepared Samples
Catalysts 2020, 10, x FOR PEER REVIEW 4 of 14 molecules plus the dehydration to Cr2O3.H2O (=2CrOOH). The second step is shown to maximize at 415 °C, giving rise to mass loss of about 11%, which is very close to the dehydration of Cr2O3.H2O into Cr2O3 (10.6%). These results verify the chemical composition of the parent chromium hydroxide compound and emphasize the appropriateness of the calcination temperature regime (≥400 °C), applied for the thermal genesis of the Cr2O3 NPs from the parent hydroxide. Accordingly, we chose the calcination temperature of 600 °C.  Figure 3 compares XRD powder diffractograms measured for Cr2O3 NPs and the hybrids Cr2O3/rGO and Cr2O3/GO. In the case of Cr2O3, the monitored diffraction peaks are quite similar to the peaks filed for α-Cr2O3 in JCPDS card no. 38-1479 [27]. Hence, α-Cr2O3 is the sole XRD-detectable crystalline phase in the calcination products at 600 °C. The average crystallite size determined for the α-Cr2O3 particles was 32 nm. In the case of Cr2O3/GO (Figure 3), the peak at around 2θ = 10.6° is assigned to the graphene oxide nanosheet, whereas the rest of the peaks monitored are indexed to the chromia NPs. As well as this, the XRD data of Cr2O3/rGO ( Figure 3) is related to JCPDS no. 38-1479, and the diffraction peak at 2θ = 24.8° could be assigned to the rGO nanosheet. It is obvious that the peak at 2θ = 10.6° has disappeared, confirming the quantitative reduction of GO to rGO [28]. Hence, the XRD patterns of the synthesized hybrids Cr2O3/rGO and Cr2O3/GO reveal their effective preparation.   [27]. Hence, α-Cr 2 O 3 is the sole XRD-detectable crystalline phase in the calcination products at 600 • C. The average crystallite size determined for the α-Cr 2 O 3 particles was 32 nm. In the case of Cr 2 O 3 /GO (Figure 3), the peak at around 2θ = 10.6 • is assigned to the graphene oxide nanosheet, whereas the rest of the peaks monitored are indexed to the chromia NPs. As well as this, the XRD data of Cr 2 O 3 /rGO ( Figure 3) is related to JCPDS no. 38-1479, and the diffraction peak at 2θ = 24.8 • could be assigned to the rGO nanosheet. It is obvious that the peak at 2θ = 10.6 • has disappeared, confirming the quantitative reduction of GO to rGO [28]. Hence, the XRD patterns of the synthesized hybrids Cr 2 O 3 /rGO and Cr 2 O 3 /GO reveal their effective preparation.  TEM image obtained for α-Cr2O3 NPs yielded after calcination at 600 °C is shown in Figure 4 A. It displayed well defined cubic-shaped nanoparticles of chromia. The average particle size for the obtained Cr2O3 NPs at 600 °C is 47 nm, which is higher than the XRD-determined average crystallite size (32 nm). This may reveal that the Cr2O3 NPs are polycrystalline and are formed via coalescence and/or aggregation of the primary crystallites. Figure 4B,C show the TEM images of the Cr2O3/rGO and Cr2O3/GO hybrids, respectively. The images clearly reveal the distribution of relatively nonaggregating chromia NPs on the GO and rGO surfaces. The HRTEM image for Cr2O3 calcined at 600 °C is shown in Figure 4D. The image shows well-resolved lattice fringes with d-spacings of 0.264 nm for rhombohedral crystallites. Accordingly, the surface-abundant crystallite facet is 104 [23]. These findings are in good agreement with the XRD results, which indicate that the (104)-facet is the strongest reflecting crystal plane, with a d-spacing value of 2.61 (in Å ).
XPS spectra for Cr(2p) and O(1s) obtained for Cr2O3 NPs after calcination at 600 °C are shown in Figure 5A,B. Cr(2p) electron BE values monitored at 586.1 and 587.8 eV are attributed to Cr 2p1/2, and the values monitored at 576.5 and 578.6 eV are assigned to Cr 2p3/2, respectively [29]. The lower BE of 576.5 eV corresponds to the Cr 3+ state [30]. The surface is also partially oxidized in order to form Cr >3+ ions, as previously reported by Fahim et al. [12]. The band at 578.6 eV is assigned to the higher oxidation state Cr 6+ (CrO3 species) [29]. The O(1s) spectrum monitors a broad peak deconvoluted into three subpeaks at binding energy 530, 531 and 532.4 eV. The peaks at 530 and 531 eV are assignable to lattice oxygen, and this is in accordance with previously reported work [31,32]. The peak located at around 532.4 eV was attributed to surface oxygen species and/or adsorbed surface hydroxyls [32]. TEM image obtained for α-Cr 2 O 3 NPs yielded after calcination at 600 • C is shown in Figure 4 A. It displayed well defined cubic-shaped nanoparticles of chromia. The average particle size for the obtained Cr 2 O 3 NPs at 600 • C is 47 nm, which is higher than the XRD-determined average crystallite size (32 nm). This may reveal that the Cr 2 O 3 NPs are polycrystalline and are formed via coalescence and/or aggregation of the primary crystallites. Figure 4B,C show the TEM images of the Cr 2 O 3 /rGO and Cr 2 O 3 /GO hybrids, respectively. The images clearly reveal the distribution of relatively non-aggregating chromia NPs on the GO and rGO surfaces. The HRTEM image for Cr 2 O 3 calcined at 600 • C is shown in Figure 4D. The image shows well-resolved lattice fringes with d-spacings of 0.264 nm for rhombohedral crystallites. Accordingly, the surface-abundant crystallite facet is 104 [23]. These findings are in good agreement with the XRD results, which indicate that the (104)-facet is the strongest reflecting crystal plane, with a d-spacing value of 2.61 (in Å).  XPS spectra for the hybrids Cr2O3/rGO and Cr2O3/GO illustrate the presence of the chromium, oxygen and carbon spectra of the investigated hybrids. The chromium spectrum shows the same deconvoluted peaks obtained in the case of Cr2O3 ( Figure 5C,D). In the C 1s spectrum of Cr2O3/GO ( Figure 5E) there are also three peaks at BEs of 289.1 for the O−C=O, 285.8 for C−OH, and 284.6 eV for C−C species. In the case of Cr2O3/rGO ( Figure 5F), there was one more peak at 286.2 eV, consequent to the epoxide (C−O−C) bond. The C−C bonds in the case of rGO are more intense than in GO. The intensity of the C−O peak also decreased further in the case of reduced rGO than in GO. The dramatic decrease of the peak's intensity in the oxygen species and the sharp increase of the C-C peak at 284.6 eV in the case of Cr2O3/rGO hybrid support the reduction process of GO to rGO. As well as this, the decrease in the Cr/C atomic ratio of Cr2O3/rGO (0.26), compared with Cr2O3/GO (0. 34), is related to a higher carbon percentage in rGO than GO.
Moreover, the surface elemental analysis indicated that the carbon-to-oxygen atomic ratio (C/O) in the case of GO and rGO is 2.2 and 3.1, respectively. Increasing the C/O ratio for rGO confirms the partial removal of oxygen through the reduction process, which is in line with the XRD results.
The Brunauer-Emmett-Teller (BET) surface area of the prepared nanoparticles of Cr2O3 NPs at 600 °C and the hybrids Cr2O3/rGO and Cr2O3/GO were measured. N2 adsorption-desorption isotherms of the examined materials are displayed in Figure 6. It is obvious that the isotherm obtained for bare Cr2O3 is a Type IV isotherm with a H1 type hysteresis loop, which implies a largely mesoporous surface [33]. Comparatively, the other two isotherms (hybrid catalysts) look like Type IV with a H3 type hysteresis loop that could be attributed to ink-bottle-like pores of varying radius, often produced by agglomerates or compacts of spheroidal particles of non-uniform size. The surface area values measured for Cr2O3 and the hybrids rGO-Cr2O3 and GO-Cr2O3 were found to be 39, 112 and 144 m 2 /g, respectively. Moreover, the corresponding mean pore volumes therefrom derived are 0.26, 0.64 and 0.93 cm 3 g −1 , respectively. XPS spectra for Cr(2p) and O(1s) obtained for Cr 2 O 3 NPs after calcination at 600 • C are shown in Figure 5A,B. Cr(2p) electron BE values monitored at 586.1 and 587.8 eV are attributed to Cr 2p 1/2 , and the values monitored at 576.5 and 578.6 eV are assigned to Cr 2p 3/2 , respectively [29]. The lower BE of 576.5 eV corresponds to the Cr 3+ state [30]. The surface is also partially oxidized in order to form Cr >3+ ions, as previously reported by Fahim et al. [12]. The band at 578.6 eV is assigned to the higher oxidation state Cr 6+ (CrO 3 species) [29]. The O(1s) spectrum monitors a broad peak deconvoluted into three subpeaks at binding energy 530, 531 and 532.4 eV. The peaks at 530 and 531 eV are assignable to lattice oxygen, and this is in accordance with previously reported work [31,32]. The peak located at around 532.4 eV was attributed to surface oxygen species and/or adsorbed surface hydroxyls [32].    Figure 7 demonstrates the catalytic activities for CO oxidation on the Cr2O3 NPs alone and the hybrid catalysts Cr2O3/rGO and Cr2O3/GO in the temperature range of 30−240 °C. T100 is the temperature at which 100% of CO has been oxidized, while T50 is the light-off temperature. As illustrated in Figure 7, the conversion percentage of CO increased with the rise of the catalytic reaction temperature. The light-off temperature of CO oxidation (T50) was achieved at 55, 98 and 115 °C in the presence of Cr2O3/GO, Cr2O3/rGO and Cr2O3, respectively. The Cr2O3/GO hybrid nanocatalyst showed the best complete CO conversion (T100) values at 120 °C. Moreover, the surface elemental analysis indicated that the carbon-to-oxygen atomic ratio (C/O) in the case of GO and rGO is 2.2 and 3.1, respectively. Increasing the C/O ratio for rGO confirms the partial removal of oxygen through the reduction process, which is in line with the XRD results.

Catalytic Activity
The Brunauer-Emmett-Teller (BET) surface area of the prepared nanoparticles of Cr 2 O 3 NPs at 600 • C and the hybrids Cr 2 O 3 /rGO and Cr 2 O 3 /GO were measured. N 2 adsorption-desorption isotherms of the examined materials are displayed in Figure 6. It is obvious that the isotherm obtained for bare Cr 2 O 3 is a Type IV isotherm with a H1 type hysteresis loop, which implies a largely mesoporous surface [33]. Comparatively, the other two isotherms (hybrid catalysts) look like Type IV with a H3 type hysteresis loop that could be attributed to ink-bottle-like pores of varying radius, often produced by agglomerates or compacts of spheroidal particles of non-uniform size. The surface area values measured for Cr 2 O 3 and the hybrids rGO-Cr 2 O 3 and GO-Cr 2 O 3 were found to be 39, 112 and 144 m 2 /g, respectively. Moreover, the corresponding mean pore volumes therefrom derived are 0.26, 0.64 and 0.93 cm 3 g −1 , respectively.    Figure 7 demonstrates the catalytic activities for CO oxidation on the Cr2O3 NPs alone and the hybrid catalysts Cr2O3/rGO and Cr2O3/GO in the temperature range of 30−240 °C. T100 is the temperature at which 100% of CO has been oxidized, while T50 is the light-off temperature. As illustrated in Figure 7, the conversion percentage of CO increased with the rise of the catalytic reaction temperature. The light-off temperature of CO oxidation (T50) was achieved at 55, 98 and 115 °C in the presence of Cr2O3/GO, Cr2O3/rGO and Cr2O3, respectively. The Cr2O3/GO hybrid nanocatalyst showed the best complete CO conversion (T100) values at 120 °C. The catalytic activity was attributed to higher surface area compared with the other catalysts and the mixed oxidation state of chromium (Cr 3+ and Cr >3+ ), which facilitates the electron mobility required for CO oxidation along with graphene oxide support. In Figure 7, the hybrid catalyst containing GO shows the highest CO conversion percent to CO2 due to the different functional groups present on GO compared with rGO, which promote the CO oxidation process. The Cr2O3/GO hybrid nanocatalysts used in this study showed much higher catalytic activity than in previous reports, as shown in Table 1. The turnover frequency (TOF) for the investigated catalysts was calculated in terms of μmol of CO converted per g of the catalyst at 60 °C to evaluate the intrinsic activity. The TOF of Cr2O3 NPs and the TOFs of Cr2O3/rGO and Cr2O3/GO nanocatalysts were found to be 11, 15 and 24 μmol g −1 min −1 , respectively. According to the results, Cr2O3/GO exhibits the best intrinsic activity, which could be related to high surface area and the presence of redox couples (Cr 3+ /Cr >+3 ) on chromia surfaces. These are important catalytic oxidation sites that facilitate the electron mobility required for The catalytic activity was attributed to higher surface area compared with the other catalysts and the mixed oxidation state of chromium (Cr 3+ and Cr >3+ ), which facilitates the electron mobility required for CO oxidation along with graphene oxide support. In Figure 7, the hybrid catalyst containing GO shows the highest CO conversion percent to CO 2 due to the different functional groups present on GO compared with rGO, which promote the CO oxidation process. The Cr 2 O 3 /GO hybrid nanocatalysts used in this study showed much higher catalytic activity than in previous reports, as shown in Table 1. The turnover frequency (TOF) for the investigated catalysts was calculated in terms of µmol of CO converted per g of the catalyst at 60 • C to evaluate the intrinsic activity. The TOF of Cr 2 O 3 NPs and the TOFs of Cr 2 O 3 /rGO and Cr 2 O 3 /GO nanocatalysts were found to be 11, 15 and 24 µmol g −1 min −1 , respectively. According to the results, Cr 2 O 3 /GO exhibits the best intrinsic activity, which could be related to high surface area and the presence of redox couples (Cr 3+ /Cr >+3 ) on chromia surfaces. These are important catalytic oxidation sites that facilitate the electron mobility required for CO oxidation.

Catalytic Activity
In addition, the Cr 2 O 3 /GO hybrid catalyst revealed good stability at 240 • C compared with pure Cr 2 O 3, which started to lose its conversion rate after 38 h and decreased up to 73% after 72 h, as shown in Figure 8. The pronounced catalytic activity of Cr 2 O 3 /GO with respect to Cr 2 O 3 NP s is believed to be due to its higher surface area of 144 m 2 /g compared with 39 m 2 /g in the case of the pure Cr 2 O 3 NPs. The decreased conversion rate is suggested to be a result of the aggregation of the pure Cr 2 O 3 NPs and the reduction of active oxygen on its surface. The proposed possible CO oxidation pathways over the Cr2O3/GO hybrid nanocatalyst are as follows: the molecular oxygen is dissociated into reactive oxygen species, and the adsorbed CO binds with active oxygen to form CO2 on the surface of Cr2O3 NPs. XPS results confirmed the presence of the redox couples (Cr 3+ /Cr >+3 ) on chromia surfaces, which is significant for the required electron mobility for CO oxidation. The adsorbed CO molecules could be extracting the surface lattice oxygen from Cr2O3 NPs to form CO2 along with surface oxygen vacancy. The CO2 molecules break away from the Cr2O3 surface, and the vacancy is filled with oxygen gas, which could be reacted with CO to form bidentate carbonate, which can then form CO2 again. Moreover, the oxygen vacancy could be filled by the active oxygen species passed through the graphene oxide. In conclusion, compared with Cr2O3, the Cr2O3/GO nanocatalyst shows the highest activity for CO oxidation due to the presence of graphene oxide substrate, which prevents the aggregation of Cr2O3 NPs and delivers more active sites. Its higher activity compared with Cr2O3/rGO is believed to be because of its higher surface area.  In the case of the hybrid catalysts, Cr 2 O 3 is dispersed well in the carbonaceous material, which prevents its aggregation, therefore maintaining its stability at the conversion rate for a long time.
The proposed possible CO oxidation pathways over the Cr 2 O 3 /GO hybrid nanocatalyst are as follows: the molecular oxygen is dissociated into reactive oxygen species, and the adsorbed CO binds with active oxygen to form CO 2 on the surface of Cr 2 O 3 NPs. XPS results confirmed the presence of the redox couples (Cr 3+ /Cr >+3 ) on chromia surfaces, which is significant for the required electron mobility for CO oxidation. The adsorbed CO molecules could be extracting the surface lattice oxygen from Cr 2 O 3 NPs to form CO 2 along with surface oxygen vacancy. The CO 2 molecules break away from the Cr 2 O 3 surface, and the vacancy is filled with oxygen gas, which could be reacted with CO to form bidentate carbonate, which can then form CO 2 again. Moreover, the oxygen vacancy could be filled by the active oxygen species passed through the graphene oxide. In conclusion, compared with Cr 2 O 3, the Cr 2 O 3 /GO nanocatalyst shows the highest activity for CO oxidation due to the presence of graphene oxide substrate, which prevents the aggregation of Cr 2 O 3 NPs and delivers more active sites. Its higher activity compared with Cr 2 O 3 /rGO is believed to be because of its higher surface area.

Preparation of GO and rGO
GO was synthesized according to the modified Hummer's method [36], wherein a 2 g portion of graphite powder was placed in a 250 mL beaker, then 1.0 g of NaNO 3 and 46 mL of sulfuric acid were added with vigorous stirring in an ice bath at 0 • C.Then, 6 g of potassium permanganate was placed slowly onto the mixture at <20 • C. After 5 min we removed the ice bath, followed by heating out at 35 • C for 30 min. Subsequently, a 92 mL aliquot of deionized water was poured slowly onto the mixture while increasing the temperature to 98 • C and stirring continuously. This was maintained for another 30 min. Subsequently, a 280 mL portion of hot water (60 • C) and 30 mL of hydrogen peroxide (H 2 O 2 ) aqueous solution (30%) were added until there was no bubble evolution. Eventually, the mixture was centrifuged for 10 min (7200 rpm) with an aqueous solution of HCl (5%) to remove excess of manganese salt; the final powder was then washed using warm water until a reddish-brown color was obtained from the black color. Finally, GO was ultrasonically dispersed in water to form a stable solution of GO.
For the synthesis of rGO, 0.1 g of GO was sonicated in 20 mL H 2 O, followed by adding 100 mL of hydrazine hydrate and heating it using a microwave at 1000 W in two 30 s cycles. A black color was obtained, confirming the complete reduction of GO to rGO [22].

Preparation of Bare Chromia Nanoparticles (Cr 2 O 3 NPs)
Cr 2 O 3 NPs were synthesized as follows: chromium nitrate was mixed with NaOH solution by stirring at 100 • C for one hour in order to obtain a dark green precipitate of Cr(OH) 3 . The precipitate was centrifuged and then washed with H 2 O to remove the NaNO 3 . Finally, the precipitate was dried at 110 • C overnight, accompanied by calcination in air at 600 • C for 5 h to obtain Cr 2 O 3 NPs.

Preparation of the Hybrid Catalysts (Cr 2 O 3 /GO and Cr 2 O 3 /rGO)
In a 100 mL round bottom flask, 0.25 g of GO or rGO was dispersed in 75 mL of THF. Subsequently, 1.0 g of the prepared Cr 2 O 3 NPs was added to the suspension and stirred continuously for 30 min. Then, the mixture was refluxed for 24 h at 70 • C, accompanied by centrifuge and washing using THF. Finally, the prepared hybrids were thermally treated at 150 • C for 4 h in a hot air oven before use in order to improve crystallinity. The hybrid materials were prepared while maintaining the support material at 25%, corresponding to Cr 2 O 3 NPs.

Characterization of Chromia NPs and Hybrid Catalysts
XPS measurements were performed by a Thermo ESCA Lab 250xi (Waltham, MA, USA) equipped with Mg Kα radiation (1253 eV). The XRD analysis was performed by a D8 Advance diffractometer (Leipzig, Germany) with a Cu target and nickel filter with CuKα radiation (λ = 0.154056 nm). Crystallite size was realized by adopting the line broadening technique and Scherrer's formula [37]. Raman spectra of GO and rGO were measured using RENISHAW (Oxfordshire, United Kingdom) via a Raman microscope, with an excitation wavelength of 514 nm, a line of an Argon ion laser, and an exposure time of 10 s. TEM data were recorded using a JEOL JEM 1230 microscope (Tokyo, Japan) at 120 kV. HRTEM was performed on a JEM-3010 JEOL ultra high-resolution electron microscope (Tokyo, Japan) operated at 300 kV and 119 µA. FESEM was performed using SEM model (JSM-6300 JEOL, Tokyo, Japan). TG was performed by heating 10 mg of test material at 10 • C/min up to 800 • C in a dynamic (50 cm 3 /min) atmosphere of air, using a model TA-50 Shimadzu thermal system (Kyoto, Japan) equipped with a workstation. N 2 adsorption-desorption isotherms were measured on test samples at liquid nitrogen temperature (−195 • C), using an automatic Micromeritics sorptiometer model ASAP 2010 (Norcross, Georgia, USA) equipped with a degassing platform and online data acquisition and handling system powered with BET-based analytical software for surface area (S BET /m 2 g −1 ) determination.

Catalytic Activity Measurements
The catalytic activity of the investigated catalysts for CO oxidation was tested using a tubular reactor adopting the same method previously reported [32]. The test catalyst powder was weighed out to 300 mg and placed on a G1-porous quartz disc mounted in the middle of a tubular reactor with an inner diameter of 2 cm and a length of 15 cm. The catalyst-bed aspect ratio (length to diameter) is 2 with short catalyst bed length of 1.3 cm. A 300 Torr portion of 1CO + 3O 2 was expanded in the reactor at room temperature (RT), followed by a temperature increase in the range of 50-400 • C. 50 µL gas samples were withdrawn at different temperatures for analysis by a CP-GC6890 gas chromatograph