Non-noble, efficient catalyst of unsupported α-Cr2O3 nanoparticles for low temperature CO Oxidation

Herein, we report the synthesis of chromium oxide nanoparticles, α -Cr2O3 NPs, followed by full characterization via XRD, SEM, XPS, and N2 sorptiometry. The synthesized nanoparticles were tested as catalysts toward the oxidation of CO. The impact of calcination temperature on the catalytic activity was also investigated. CO conversion (%), light-off temperature, T50, data were determined. The results revealed that chromia obtained at low calcination temperature (400 °C) is more active than those obtained at high calcination temperatures (600° or 800 °C) and this is ascribed to the smaller particle size and higher surface area of this sample. The results revealed a superior catalytic activity of Cr2O3 NPs at lower temperature as we reached a complete conversion at 200 °C which is high value in the forefront of the published results of other non-noble catalysts. The high activity of Cr2O3 nanoparticles (T50 as low as 98 °C) where found to be dependent on a careful selection of the calcination temperature. These results may provide effective and economic solutions to overcome one of the major environmental threats.


Experimental part
Preparation of chromia nanoparticles. 25% aqueous ammonia solution was added dropwise to a continuously stirred 0.1 M aqueous solution of the nitrate salt of Cr. The resulting mixture was stirred for another hour, then was left overnight. The precipitate thus obtained was filtrated through a Wattman filter paper (No. 42), washed with double distilled water, and dried overnight at 100 °C. The dried precipitate was ground, sealed in vials, and stored over silica gel in a desiccator till further use. The dried precipitates were treated thermally by calcination on heating in a still atmosphere of air at various temperatures (400, 600, or 800 °C) for 2 h. The calcination products thus obtained are indicated below by corresponding oxide formula and an added Arabic numeral to symbolize the temperature applied.
Characterization of Nanoparticles. The X-ray diffraction (XRD) measurements were conducted by using a Bruker AXS D8 Advance X-Ray Powder Diffractometer with a copper target and a nickel filter with Cu Kα radiation (λ = 0.154 nm). Measurements were performed in the range 20-80° (2θ). The morphology of the particles, as well as electron diffraction patterns were obtained by scanning electron microscopy (SEM) using a JEOL JSM-7001F operating at 120 kV.
X-ray photoelectron spectroscopy (XPS) surface elemental analysis was conducted using a model Thermo ESCA Lab 250xi equipped with Mg Kα radiation (1253 eV) and operated at 23 kV and 13 mA. The binding energy was referenced to the C 1s line at 284.76 eV for calibration. N 2 adsorption-desorption isotherms were measured on test samples at liquid nitrogen temperature (−195 °C) using a model ASAP 2010 automatic Micromeritics sorptiometer (USA) equipped with a degassing platform.
Catalytic activity measurements. A weighed portion (250-300 mg) of the test catalyst, was placed at ambient temperature on a G1-porous quartz disc mounted in the middle of a tubular reactor (i.d. = 2 cm; length = 15 cm) equipped with a tubular, sealed sheath for a Pt/Rh thermocouple compatible with a Type-J Cole-Parmer (USA) digital thermometer for reaction temperature reading accurate to within ±2 °C. The catalyst was pre-activated by in situ heating (at 20 °C min −1 ) in a 200 Torr portion of O 2 gas at 200 °C for 15 min. The heating was enabled by a temperature-controlled OMEGA (USA) tubular furnace mounted around the reactor. Subsequently, the gas was pumped off to 10 −2 Torr, and a fresh 200 Torr portion of O 2 was admitted into the reactor and maintained for 15 min, prior to further outgassing. This process of gassing and outgassing of O 2 at 200 °C was repeated for two more times, before a final outgassing at 200 °C and cooling to room temperature (RT). A 300 Torr portion of the reactants mixture (CO + O 2 with a 1:3 mass ratio) was expanded into the reactor at RT, then temperature was increased at 20 °C min −1 to certain higher temperatures in the range 50-400 °C keeping an almost constant interval of ca. 50 °C. Each temperature was maintained for a 15-min period through which the gas samples were withdrawn from the reaction mixture (reactants plus products). It is worth noting, that the reactor hot zone was counter-parted by a cold zone (a refrigerated water jacket) maintained at 10 °C, adopting a design similar to that described previously by Schwab et al. 21 .
50-μl gas samples were withdrawn from the reaction atmosphere of catalytic CO oxidation by means of a Hamilton gas syringe at different reaction temperatures (three samples at each temperature) and analyzed by a model CP-9001 Chrompack gas chromatograph (The Netherlands). The GC was equipped with a TCD detector (maintained at 120 °C) and a packed column of PORAPACK Q (maintained at 110 °C). A 99.99% pure He (KOAC) was employed as the carrier gas (20 ml min −1 ), and an installed MOSAIC software facilitated a computer processing of the experimental results versus pre-constructed calibration curves. For accurate quantitative results, calibration procedures are required. Calibration involves the correlation between a known concentration of a component and the resultant detector signal generated when that component is detected. Electronic integrators are used to convert the detector signal to either peak area or height. Calibration algorithms are incorporated into the integrator or computer software used for data acquisition and analysis. Thus, one creates calibration tables in the data system through the use of prompted dialogs or menu items. The simplest procedure involves creating a single-level calibration and is achieved by the following process. First one optimizes the chromatographic parameters for the desired separation and identification of components (such as choice of column, detector, injection process, and oven temperature).

Results and Discussions
Characterization of chromia nanoparticles. IR spectra obtained for the as-obtained chromia, Cr 2 O 3 -RT, and its calcination products at 400-800 °C are shown in Fig. 1. IR spectrum of the Cr 2 O 3 sample displays a strong absorption at 523 cm −1 , a broad shoulder at 839 cm −1 , a sharp band at 1384 cm −1 , a weak shoulder at 1468 cm −1 and a weak band 1628 cm −1 . The spectrum also exhibits a strong and broad band centered around 3416 cm −1 . The broad band shown at 3416 cm −1 is due to υO-H of H-bonded hydroxyl groups. The peak at 1628 cm −1 can be assigned to δH-O-H bending vibration of physically adsorbed water molecule. The band at 1467 cm −1 is attributed to υCOO of carbonate impurities species. The absorption at 1384 cm −1 is due to υNO 3 − vibrations, which originate from the metal precursor. The bands below 1000 cm −1 is related to the υCr-O bands. IR spectra of the calcination products (at 400-800 °C), are significantly different from that of the as-prepared Cr 2 O 3 . The IR spectrum of Cr 2 O 3 -4 only exhibits two sharp, strong bands at 640 and 581 cm −1 , in addition to a much weaker absorption at 3453 cm −1 . Upon further increase of the calcinations temperature up to 800 °C two additional low-frequency bands emerge at 445 and 414 cm −1 . The four absorption at 661, 574, 445 and 414 cm −1 are assignable to υCr−O lattice vibration of α-Cr 2 O 3 crystallites 22 .
XRD diffractogram recorded for the as-obtained Cr 2 O 3 and its calcination products (400-800 °C) are exhibited in Fig. 2. It is obvious from the results that the as-obtained Cr 2 O 3 is largely non-crystallite, which is reflected in its gelatinous nature. The calcination products are shown to be crystalline. The crystallinity is shown to increase with calcination temperature. In spite of that, the calcination products are shown to exhibit the same diffraction Scientific RepORts | 7: 14788 | DOI:10.1038/s41598-017-14779-x pattern, which is similar to that filed in JCPDS card no. 01-1294 23 for α-Cr 2 O 3 . The crystal size can be calculated according to Debye-Scherrer formula: where K, a shape factor, k = 0.89, λ is the wavelength of the Cu-Kα radiations, ß is the full width at half maximum and θ is the angle obtained from 2θ values corresponding to maximum intensity peak in XRD pattern. The mean crystallite size values were found to be an increase function of temperature and equal 36 nm, 40 nm and 56 nm for Cr 2 O 3 -4, Cr 2 O 3 -6 and Cr 2 O 3 -8 respectively. XPS results obtained for a chromia, Cr 2 O 3 -4, are shown in Fig. 3. The binding energy of all the deconvoluted peaks of the studied samples are tabulated in Table 1. Cr(2p) electron binding energy values were found to be located at 586 and 587ev for Cr 2p 1/2 , and 576 and 579 eV for Cr 2p 3/2 respectively. The lower binding energy peak located at 576 eV is assigned to the Cr 3+ state 24 . While the peak located at 579 eV is attributed to higher oxidation state Cr 6+ 24 . The results indicate that despite the fact that the calcination products are essentially α-Cr 2 O 3 , the surface is partially oxidized to expose Cr ≥3+ ions. Similar results have been reported by Fahim, et al. 25 . The Cr ≥3+ / Cr 3+ surface atomic ratios, as were calculated from 2p XPS signal of Cr, are 0.18, 0.32, 0.42, 0.31 for Cr 2 O 3 -RT, Cr 2 O 3 -4, Cr 2 O 3 -6, and Cr 2 O 3 -8, respectively. The slight amount of surface OH -1 is due to the unavoidable physical water adsorption at the surface 26 . The O1s XPS spectrum showed three components with O1s binding energies at 529.7, and 531.1 eV, assigned to lattice oxygen and the peak at 533.2 eV is assigned to surface defects and/or adsorbed surface hydroxyls 27 . The O/Cr surface atomic ratio were found to be 4.6, 2.8, 2.1, and 2.1 for Cr 2 O 3 -RT,  The higher ratio at low calcination temperature is expected and it is due to the formation of Cr 2 O 3 .nH 2 O structure (oxide-hydroxide structure). This is also evident from the presence of two type of O 2− signals shows in Table 1, the lower binding energy one's is believed to be for surface lattice oxygen from the nonhydrated Cr 2 O 3 formula while the higher bending energy is from water associated with the chromia formula (i.e. hydrated Cr 2 O 3 ). The third peak at around 533 eV are corresponding to the OH − /H 2 O species physisorbed at the surface. In other words, beside the XPS signal of the lattice oxygen, there is two other signals resulted from two type of water interacting with the catalyst surfaces. One resulted from chemisorbed water encouraged by the presence of lattice defects/vacancies and the other is due to physisorbed water molecules experiencing dipole-dipole interactions with the lattice oxygen.   Fig. 4. It is obvious from the classification reviewed elsewhere 28 , the isotherms obtained are of type-IV. Type-IV isotherms imply largely mesoporous surfaces. Except for the as-obtained α-Cr 2 O 3 , i.e., non-crystalline chromia, the hysteresis loop displayed has closure points at p/p • < 0.45 which may account for narrow mesopores. Hysteresis loops exhibited by as-obtained α-Cr 2 O 3 and its calcination products (400°-600 °C) are of type-H2. Whereas α-Cr 2 O 3 -8 displays type-H3 loop. Type-H2 loop is believed to be associated with ink-bottle-like pores of varying radius, often generated by agglomerates or compacts of spheroidal particles of non-uniform size and arrangement. A type-H3 loop is attributed to adsorbate condensation in capillary spaces between parallel plates or open slit-shaped capillaries.
The S B values observed for the as-obtained Cr 2 O 3 is 127 m 2 /g. Upon calcination at 400°−800 °C, the S B value decreased to 29 and further to 17 m 2 /g (see Table 2). A S B ≈ S C indicates presence of cylindrical mesopores while the result (S C > S B ) exhibited by Cr 2 O 3 -8 implies the presence of non-cylinderical mesopores. The Cr 2 O 3 showed lowest pore volume value (0.0385 cm 3 /g) at 800 °C (Table 2), which is a direct result of the high temperature calcination and consequent sintering of particles. The as-obtained Cr 2 O 3 itself shows the highest pore volume (0.235 cm 3 /g). Upon calcination (400°-800 °C) α-Cr 2 O 3 particles are produced, which exhibit a higher V c p compared to V p thus implying the presence of non-cylindrical mesopores (see Table 1). The as-obtained α-Cr 2 O 3 shows narrower pore size than the calcined ones. As the calcination temperature increases from 400° up to 800 °C, a pore widening is observed (r p /nm = 11 → 23 nm). Concomitantly, a drop in pore volume (0.0937 → 0.0385 cm 3 /g) is observed. This supports a pore widening mechanism ( Table 2). SEM images obtained for as-obtained Cr 2 O 3 and its calcination products at 400°−800 °C are shown in Fig. 5. SEM indicates that the as-obtained Cr 2 O 3 and Cr 2 O 3 -4 products exhibit irregular clumps of particles of ill-defined contours, whereas the calcined samples at ≥600°C produce a different morphology characterized by uniform particle size showing quazi-spherical particles. It is also seen that the size of these particles increases with calcination emperature from 600° to 800 °C.
CO oxidation activity. On chromia surfaces, redox couples of Cr 3+ /Cr >+3 are important catalytic oxidation sites 29 . Figure 6A compares plots of CO conversion versus reaction temperature for chromia catalysts derived from Cr 2 O 3 at various calcination temperatures.
The results brought about reveal that CO oxidation on the chromia catalysts occurs actively at low temperatures as shown in Fig. 6A. T 100 is the temperature at which 100% of CO was oxidized and T 50 is the light-off temperature and their corresponding values are represented in Table 3 in which the present results were compared   Table 2. BET surface area (S B /m 2 /g), -specific surface area (S C /m 2 /g), pore volume (V p (cm 3 /g) and radius (r p (nm) and crystallite size (L/ ± 1 nm), for the as-obtained Cr 2 O 3 and its calcination products at 400°-800 °C.

V P /(cm 3 /g) r p /nm S B /(m 2 /g) S C /(m 2 /g) L/nm
with their corresponding non-noble catalysts reported in the literature. Our results also reveal that chromia obtained at low calcination temperature (400°C) is more active than those obtained at high calcination temperatures (600° or 800 °C). From the XPS data, we found that, the higher the calcination temperature, the lower the surface content of Cr >+3 sites. Since it is known that on chromia surfaces intimately coupled Cr 3+ -Cr >+3 sites facilitate the electron mobility required for CO oxidation, and, thus, considered the active sites 29 . Hence, the drop in activity upon increasing calcination temperature may be linked to the concomitant loss or excess changes on chromia surfaces. It is worth noting, that the three test samples of chromia are of comparable surface area, thus the surface area may not be a key parameter in shaping up the CO oxidation activity on the present test samples.  Moreover, chromia is an acidic oxide, and, hence, surface basic sites are also of less importance to the surface catalytic activity in the present case. The results presented in Fig. 6B, are for the intrinsic (i.e. exclude the specific surface area from consideration) activity of the calcined catalyst. In other words, the graphs show the turn over number, TON, per CO molecules per min per m 2 . Accordingly, TON-based ranking of the test oxides shows that at 100 °C, the Cr 2 O 3 -4 samples exhibit a better intrinsic activity than Cr 2 O 3 -6 and Cr 2 O 3 -8. This means that, among the studied catalysts, Cr 2 O 3 -4 exhibits the best activity and 100 °C intrinsic activity. It is worth mentioning that, from the XPS results, the Cr ≥3+ / Cr 3+ surface atomic ratios of Cr 2 O 3 -4 is 0.32 which is equal to that found for Cr 2 O 3 -8 (0.31) but less than that of Cr 2 O 3 -6 (0.42). This means that, for the studied samples, the Cr ≥3+ /Cr 3+ surface ratio is not the only factor that control the intrinsic activity and other factors are playing roles. This also support our conclusion that the occurrence of Cr 2 O 3 -4 at the transition intermediate between Cr 2 O 3 .nH 2 O ⇔ Cr 2 O 3 is the determining factor here. The different morphological structure of the Cr 2 O 3 -4 catalyst (see Fig. 5) may also play role in such superiority 30 .
Kinetic catalysis studies have shown the pathway adopted by CO oxidation on metal oxides to be dependent on the reaction temperature regime applied 31 : (i) at ≤150 °C, Eley-Rideal mechanism (CO (g) + O 2 (ads) or O 2(g) + CO(ads), ∆E ≤ 10 kCal/mol; (ii) at 150-250 °C, Langmuir-Hinshelwood mechanism (CO(ads) + O 2 (ads)), ∆E ca. 10-20 kCal/mol; and (iii) at ≥250 °C, Mars-van-Krevelen mechanism (CO (g) + O(lattice)), ∆E ≥ 20 kCal/ mol. Based on the studied temperature, the Langmuir-Hinshelwood mechanism is expected (both the reactants (CO and O 2 ) are adsorbed on the catalyst surface and O 2 get activeated and reacted with the CO. The involvement of lattice oxygen is rolled out since 100% CO conversion is reach at relatively low temperature (200-300 °C). The four elementary steps of the Langmuir-Hinshelwood mechanism of CO oxidation. with the following elementary reaction Eqs (2-5): where "X" denotes the active site.

Conclusion
In this study, we reported a new CO oxidation catalyst of Cr 2 O 3 nanoparticles prepared via hydrothermal technique. The effect of calcination temperature on the crystal structure and composition was investigated via XRD  . We attributed this superior activity to this chemical composition nature in addition to its other surface characteristics. The results of this study will make its applicability to be economically possible in automotive exhausts and factory chimneys.