Synthesis of bulk and alumina-supported γ-Mo2N catalysts by a single-step complex decomposition method

A single-step complex decomposition method for the synthesis of bulk and alumina-supported γ-Mo₂N catalysts is described. The complex precursor (HMT)₂(NH₄)₄Mo₇O₂₄·2H₂O (HMT: hexamethylenetetramine) is converted to γ-Mo₂N under a flow of Ar in a temperature range of 823–1023 K. Furthermore, decomposition of the precursor in a NH₃ flow forms γ-Mo₂N in a temperature range of 723–923 K. Compared with direct decomposition of the precursor in Ar, the reaction in NH₃ shows obvious advantages that the nitride forms at a lower temperature. In addition, alumina-supported γ-Mo₂N catalysts with different nitride loadings can be prepared from the alumina-supported complex precursor in the Ar or NH₃ flow. The resultant catalysts exhibit good dibenzothiophene HDS activities, which are similar to the γ-Mo₂N/γ-Al₂O₃ prepared by traditional TPR method. The catalyst prepared by decomposition in an Ar flow exhibits highest activity. It proves that such a single-step complex decomposition method possesses the potential to be a general route for the preparation of molybdenum nitride catalysts.     

Mesoporous molecular sieve MCM-41 supported Co–Mo catalyst for hydrodesulfurization of petroleum resids

In this work, we explored the potential of mesoporous zeolite-supported Co–Mo catalyst for hydrodesulfurization of petroleum resids, atmospheric and vacuum resids at 350–450°C under 6.9 MPa of H₂ pressure. A mesoporous molecular sieve of MCM-41 type was synthesized; which has SiO₂/Al₂O₃ ratio of about 41. MCM-41 supported Co–Mo catalyst was prepared by co-impregnation of Co(NO₃)₂·6H₂O and (NH₄)₆Mo₇O₂₄ followed by calcination and sulfidation. Commercial Al₂O₃ supported Co–Mo (criterion 344TL) and dispersed ammonium tetrathiomolybdate (ATTM) were also tested for comparison purposes. The results indicated that Co–Mo/MCM-41(H) is active for HDS, but is not as good as commercial Co–Mo/Al₂O₃ for desulfurization of petroleum resids. It appears that the pore size of the synthesized MCM-41 (28 Å) is not large enough to convert large-sized molecules such as asphaltene present in the petroleum resids. Removing asphaltene from the resid prior to HDS has been found to improve the catalytic activity of Co–Mo/MCM-41(H). The use of ATTM is not as effective as that of Co–Mo catalysts, but is better for conversions of >540°C fraction as compared to noncatalytic runs at 400–450°C.     

Effect of synthesis conditions on the molybdenum nitride catalytic activity

Properties of the molybdenum nitrides supported on alumina as well as on active carbon has been investigated in the reactions of thiophene and vacuum gas oil (VGO) hydrodesulfurization (HDS) as well as in the reaction of cyclohexene with hydrogen. Supported molybdenum nitride was more active in thiophene hydrogenolysis than sulfided Mo/Al₂O₃. Also in the reaction of VGO HDS alumina supported molybdenum nitride was more active than sulfided counterpart, however Mo₂N supported on active carbon exhibit low activity. In the products of the cyclohexene reaction over supported Mo₂N methyl-cyclopentanes and benzene has been found.     

Hydrotreating of heavy vacuum gas oil (HVGO) on molybdenum and tungsten nitrides catalytic phases

A series of supported and unsupported Mo₂N and W₂N phases were synthesized by means of the treatment under ammonia atmosphere at 700°C of Mo and W oxides. The X-ray diffraction and electron microscopy techniques verified the formation of the Mo₂N and W₂N ceramic phases, while the N₂ adsorption (BET) was used to determine the surface areas, between 46–133 m²/g for Mo₂N (unsupported) and 81–101 m²/g for W₂N (unsupported). The supported phases had surface areas between 109–113 and 109–122 m²/g, for Mo₂N/Al₂O₃ and W₂N/Al₂O₃, respectively. The catalytic hydrotreating of a heavy vacuum gas oil (HVGO) derived from Maya crude (i.e. 2.21 wt.% S, 0.184 wt.% N₂) was performed on both, supported and unsupported Mo nitrides and W nitrides, which promoted the HDN reaction preferentially, up to 26.6% on Mo₂N/Al₂O₃ and up to 22.3% on W₂N/Al₂O₃, against 3.26% on the reference catalyst, i.e. CoMo/Al₂O₃ at 350°C and 80 kg/cm². Also, the rates for HDN increased with the crystallite size in the unsupported W₂N series. Also, the pore volume and mean pore diameters of the Mo₂N/Al₂O₃ and W₂N/Al₂O₃ series improve substantially with respect to the pure ceramic phases.     

A model catalyst approach to the effects of the support on Co–Mo hydrodesulfurization catalysts

Co–Mo model sulfide catalysts, in which CoMoS phases are selectively formed, were prepared by means of a CVD technique using Co(CO)₃NO as a precursor of Co. It is shown by means of XPS, FTIR and NO adsorption that CoMoS phases form selectively when the Mo content exceeds monolayer loading. A single exposure of MoS₂/Al₂O₃ to a vapor of Co(CO)₃NO at room temperature fills the edge sites of the MoS₂ particles. It is suggested that the maximum potential HDS activity of MoS₂/Al₂O₃ and Co–Mo/Al₂O₃ catalysts can be predicted by means of Co(CO)₃NO as a “probe” molecule. An attempt was made to determine the fate of Co(CO)₃NO adsorbed on MoS₂/Al₂O₃. The effects of the support on Co–Mo sulfide catalysts in HDS and HYD were investigated by use of CVD-Co/MoS₂/support catalysts. XPS and NO adsorption showed that model catalysts can also be prepared for SiO₂-, TiO₂- and ZrO₂-supported catalysts by means of the CVD technique. The thiophene HDS activity of CVD-Co/MoS₂/Al₂O₃, CVD-Co/MoS₂/TiO₂ and CVD-Co/MoS₂/Al₂O₃ is proportional to the amount of Co species interacting with the edge sites of MoS₂ particles or CoMoS phases. It is concluded that the support does not influence the HDS reactivity of CoMoS phases supported on TiO₂, ZrO₂ and Al₂O₃. In contrast, CoMoS phases on SiO₂ show catalytic features characteristic of CoMoS Type II. With the hydrogenation of butadiene, on the other hand, the Co species on MoS₂/TiO₂, ZrO₂ and SiO₂ have the same activity, while the Co species on MoS₂/Al₂O₃ have a higher activity.     

DRIFT study of temperature programmed desorption of NO adsorbed on Co–Mo/Al2O3 sulfided at high pressure

In order to make clear the coordinatively unsaturated sites (CUS) of Co–Mo/Al₂O₃ sulfided at high pressure, the temperature programmed desorption of NO adsorbed on Co–Mo/Al₂O₃ sulfided at high pressure was studied by DRIFT method. The intensity of two IR bands (1835 and 1785 cm⁻¹) of adsorbed NO on Co–Mo/Al₂O₃decreased simultaneously up to 393 K. The higher frequency band disappeared at 393 K, while the lower frequency band remained even at 403 K. In the case of Mo/Al₂O₃, the intensities of two bands appeared at 298 K decreased monotonously with increasing temperature, and disappeared simultaneously over 433 K. In the case of Co/Al₂O₃, two bands disappeared simultaneously over 393 K. These results suggest that two kinds of nitrosyl species are formed on Co–Mo/Al₂O₃. One is dinitrosyl species adsorbed on CUS of Co, and the other is unidentified nitrosyl species. Comparing DRIFT spectra of Co–Mo/Al₂O₃ with those of a physical mixture of Mo/Al₂O₃ and Co/Al₂O₃, it is also suggested that the formation of the latter one correlates with the interaction between Co and Mo in Co–Mo/Al₂O₃. The unidentified nitrosyl species might be the key to explain the dependency of DRIFT spectrum of adsorbed NO on the pressure of sulfiding.     

Effect of phosphorus addition on the hydrotreating activity of NiMo/Al2O3 carbide catalyst

A series of phosphorus promoted γ-Al₂O₃ supported NiMo carbide catalysts with 0–4.5 wt.% P, 13 wt.% Mo and 2.5 wt.% Ni were synthesized and characterized by elemental analysis, pulsed CO chemisorption, BET surface area measurement, X-ray diffraction, near-edge X-ray absorption fine structure, DRIFT spectroscopy of CO adsorption and H₂ temperature programmed reduction. X-ray diffraction patterns and CO uptake showed the P addition to NiMo/γ-Al₂O₃ carbide, increased the dispersion of β-Mo₂C particles. DRIFT spectra of adsorbed CO revealed that P addition to NiMo/γ-Al₂O₃ carbide catalyst not only increases the dispersion of Ni-Mo carbide phase, but also changes the nature of surface active sites. The hydrodenitrogenation (HDN) and hydrodesulfurization (HDS) activities of these P promoted NiMo/γ-Al₂O₃ carbide catalysts were performed in trickle bed reactor using light gas oil (LGO) derived from Athabasca bitumen and model feed containing quinoline and dibenzothiophene at industrial conditions. The P added NiMo/γ-Al₂O₃ carbide catalysts showed enhanced HDN activity compared to the NiMo/γ-Al₂O₃ catalysts with both the feed stocks. The P had almost no influence on the HDS activity of NiMo/γ-Al₂O₃ carbide with LGO and dibenzothiophene. P addition to NiMo/γ-Al₂O₃ carbide accelerated CN bond breaking and thus increased the HDN activity.     

The effects of fluorine, phosphate and chelating agents on hydrotreating catalysts and catalysis

The effects of fluorine, phosphate and chelating agents on hydrodesulfurization (HDS) and hydrodenitrogenation (HDN) are reviewed. All three additives enhance the activity of NiMo/Al₂O₃ catalysts in HDN but have only a slightly positive or even a negative effect on the HDS activity of CoMo/Al₂O₃ and NiMo/Al₂O₃ catalysts. The positive effect on HDN is due to the enhancement of the hydrogenation of aromatic rings. On the other hand, these three additives diminish the rates of C–N bond breaking and alkene hydrogenation reactions.

All three additives are hard basic ligands that may interact strongly with hard acids such as coordinatively unsaturated Al³⁺ cations on the alumina surface. A strong interaction with the alumina support has several effects. First, molybdate and tungstate anions are no longer strongly bonded to the support and are predominantly present as polyanions, which can be easily sulfided to MoS₂ and WS₂ crystallites. The weaker interaction with the smaller support surface also leads to larger MoS₂ and WS₂ crystallites with a lower dispersion. Second, the Ni²⁺ and Co²⁺ cations will also interact more weakly with the alumina, and this makes the formation of Ni and Co promoter atoms in the catalytically active Ni–Mo–S and Co–Mo–S phases more efficient. Third, the weaker interaction of Mo and W with the support leads to a higher stacking of the MoS₂ and WS₂ crystallites and, thus, to the more active type II Ni–Mo–S and Co–Mo–S phases. The increased stacking is beneficial for geometrically demanding reactions such as the hydrogenation of aromatics. For less demanding reactions, such as alkene hydrogenation, aliphatic C–N bond breaking and thiophene HDS, the loss in dispersion is important.     

Development of new catalysts for deep hydrodesulfurization of gas oil

TiO₂–Al₂O₃ composite supports have been prepared by chemical vapor deposition (CVD) over γ-Al₂O₃ substrate, using TiCl₄ as the precursor. High dispersion of TiO₂ overlayer on the surface of Al₂O₃ has been obtained, and no cluster formation has been detected. The catalytic behavior of Mo supported on Al₂O₃, TiO₂ and TiO₂–Al₂O₃ composite has been investigated for the hydrodesulfurization (HDS) of dibenzothiophene (DBT) and methyl-substituted DBT derivatives. The conversion over the Mo catalysts supported on TiO₂–Al₂O₃ composite, in particular for the HDS of 4,6-dimethyldibenzothiophene (4,6-DMDBT) is much higher than that of conversion obtained over Mo catalyst supported on Al₂O₃. The ratio of the corresponding cyclohexylbenzenes/biphenyls is increased over Mo catalyst supported on TiO₂–Al₂O₃ composite support. This means that the reaction rate of prehydrogenation of an aromatic ring rather than the rate of hydrogenolysis of C–S bond cleavage is accelerated for the HDS of DBT derivatives. The Mo/TiO₂–Al₂O₃ catalyst leads to higher catalytic performance for deep HDS of gas oil.     

Hydrodesulfurization of dibenzothiophenes over molybdenum catalyst supported on TiO2–Al2O3

Composite types of TiO₂–Al₂O₃ supports, which are γ-aluminas coated by titania, have been prepared by chemical vapor deposition (CVD), using TiCl₄ as a precursor. Then supported molybdenum catalysts have been prepared by an impregnation method. As supports, we employed γ-alumina, anatase types of titania, and composite types of TiO₂–Al₂O₃ with different loadings of TiO₂. We studied the conversion of Mo from oxidic to sulfidic state through sulfurization by X-ray photoelectron spectroscopy (XPS). The obtained spectra unambiguously revealed the higher reducibility from oxidic to sulfidic molybdenum species on the TiO₂ and TiO₂–Al₂O₃ supports compared to that on the Al₂O₃ support. Higher TiO₂ loadings of the TiO₂–Al₂O₃ composite support led to higher reducibility for molybdenum species. Furthermore, the catalytic behavior of supported molybdenum catalysts has been investigated for hydrodesulfurization (HDS) of dibenzothiophene (DBT) and methyl-substituted DBT derivatives. The conversion over the TiO₂–Al₂O₃ supported Mo catalysts, in particular for the 4,6-dimethyl-DBT, is much higher than that obtained over Al₂O₃ supported Mo catalyst. The ratio of the corresponding cyclohexylbenzene (CHB)/biphenyl (BP) derivatives is increased over the Mo/TiO₂–Al₂O₃. This indicates that the prehydrogenation of an aromatic ring plays an important role in the HDS of DBT derivatives over TiO₂–Al₂O₃ supported catalysts.     

In situ XAFS analysis of catalytically active cobalt species in porous clays for deep hydrodesulfurization

Mesoporous smectite-type clays containing cobalt species in lattice (MST(Co)) were prepared by a hydrothermal method and examined for hydrodesulfurization (HDS). The MST(Co) catalysts showed higher HDS activities than a commercial alumina-supported cobalt–molybdenum catalyst (Co–Mo/Al₂O₃). The active structure of MST(Co) was studied by in situ X-ray absorption fine structure and nitrogen adsorption techniques.     

TPD and XRD studies of molybdenum nitride and its activity for hydrodenitrogenation of carbazole

The nitrided MoO₃ catalysts formed using two kinds of treatment with either NH₃ or He after nitriding were studied by temperature-programmed desorption and X-ray diffraction analyses. The catalysts were cooled to room temperature in either flowing NH₃ or He (NH or HE catalyst) after nitriding at 773, 973 and 1173 K with NH₃. The activities of the catalysts were determined during the hydrodenitrogenation of carbazole at 573 K and 10.1 MPa total pressure. MoO₂, γ-Mo₂N, and Mo metal were mainly formed in the NH catalysts nitrided at 773, 973 and 1173 K, respectively. Mo oxides and metals in the NH catalysts were nitrided to γ-Mo₂N and β-Mo₂N_0.78 with low crystallinity during TPD. The surface area of the NH and HE catalysts nitrided at 773 K increased to 66 and 59 m² g⁻¹ maximum from 1.1 m² g⁻¹ of fresh MoO₃, respectively, but decreased as the nitriding temperature increased to 973 K and 1173 K. The HE catalysts per surface area were more active than the NH catalysts for both the overall HDN reaction and hydrogenation, and the 1173 K-nitrided catalysts were highest. On the other hand, the NH catalysts were more active than the HE catalysts for C–N hydrogenolysis and the 973 K-nitrided catalyst showed a maximum activity for C–N hydrogenolysis.     

还原温度对Mo基催化剂物相及其加氢脱氧性能的影响

以碳纳米管(CNTs)为载体,通过控制催化剂合成的还原温度制备了一系列负载型Mo基催化剂。采用XRD、TEM、N₂物理吸附、XPS以及NH₃/H₂-TPD等技术对催化剂进行了表征,并研究了Mo基催化剂对硬脂酸催化加氢脱氧性能的影响。结果表明:随着还原温度的升高,催化剂表面的Mo物种逐渐被还原,还原过程为:MoO₃→MoO₂→Mo→Mo₂C。还原温度为450℃和550℃时,催化剂的活性相为MoO₂;还原温度为600℃时,催化剂的活性相为MoO₂/Mo/β-Mo₂C的混合相;还原温度为650℃和700℃时,催化剂的活性相全部转化为β-Mo₂C。与活性相MoO₂催化剂相比,β-Mo₂C催化剂具有更高的加氢脱氧活性。此外,还原温度为600℃的MoO₂/Mo/β-Mo₂C混合相催化剂因具有较大的比表面积、较多的酸中心数量和较强的H₂吸附能力,使得该催化剂在硬脂酸加氢脱氧反应中表现出最优越的催化活性。     

Simultaneous hydrodesulfurization of thiophene and hydrogenation of cyclohexene over dimolybdenum nitride catalysts

A series of unsupported dimolybdenum nitride (γ-Mo₂N) catalysts differing in surface area were prepared by temperature programmed reduction of MoO₃ with a mixture of NH₃:N₂ (90:10). Characterization of catalysts by BET, XRD, TPR and XPS techniques was carried out. The samples were used as catalysts in hydrotreating reactions (simultaneous hydrodesulfurization of thiophene and hydrogenation of cyclohexene). Low surface area γ-Mo₂N materials show much higher specific conversions than those with higher surface area. These results indicate that HDS and HYD reactions over γ-Mo₂N seem to be structure-sensitive. The relative exposure extent of crystalline planes (1 1 1) and (2 0 0) over the different catalysts can be associated with their hydrogen adsorption capacities and with their catalytic performances. The catalytic activities are significantly affected by the catalyst pretreatment conditions.     

K/Ni/β-Mo2C: A highly active and selective catalyst for higher alcohols synthesis from CO hydrogenation

Nickel and potassium promoted β-Mo₂C catalysts were prepared for CO hydrogenation to higher alcohols synthesis. The results revealed that β-Mo₂C produced mainly hydrocarbons, but the addition of potassium resulted in a remarkable selectivity shift from hydrocarbons to alcohols over β-Mo₂C. Moreover, it was found that potassium enhanced the ability of chain propagation of β-Mo₂C catalyst and led to a higher selectivity to C₂⁺OH. The addition of nickel further enhanced higher alcohols synthesis, which showed the optimum at 1/8–1/6 of Ni/Mo molar ratios. The characterization suggested that there might be a synergistic effect of potassium and nickel on β-Mo₂C, which favored the alcohols synthesis. The production of alcohols appeared to be relevant to the presence of Mo⁴⁺ species, whereas the formation of hydrocarbons was closely associated with Mo²⁺ and/or Mo⁰ species on the surface of β-Mo₂C-based catalysts.     

In situ XAFS analysis of Pd–Pt catalysts during hydrotreatment of model oil

Supported Pd–Pt catalysts are efficient for hydrodesulfurization (HDS) and hydrodearomatization (HDA) reactions of diesel fuel and their activity varied with the kinds of supports. Concerning HDA, alumina supported catalysts showed four times higher TOF (turn over frequency) than silica supported one. In order to elucidate the difference in activity, the structural analysis of the active phase was performed. After reduction pretreatment, relatively uniform and large metallic alloy Pd–Pt particles were formed on SiO₂, whereas, Pd and Pt atoms formed rather segregated particles on Al₂O₃. Subsequent X-ray absorption of fine structure (XAFS) analysis under HDS conditions showed no contribution of sulfur for SiO₂ supported catalyst, whereas, formation of sulfided metal species was observed in XAFS spectra for the Al₂O₃ supported catalyst. It is suggested that on Pd–Pt/SiO₂, thin sulfide layer on the metal cluster surface blocked the active sites and lowered the HDA activity. Presence of partially sulfided phase originated from rather segregated structure like Pd–Pt/Al₂O₃ is thought to be requisite for high HDA activity.     

Support effects on structure and activity of molybdenum oxide catalysts for the oxidative dehydrogenation of ethane

The structural and catalytic properties of MoO₃ catalysts supported on ZrO₂, Al₂O₃, TiO₂ and SiO₂ with Mo surface densities, n_s, in the range of 0.5–18.5 Mo/nm² were studied for the oxidative dehydrogenation (ODH) of ethane by in situ Raman spectroscopy and catalytic activity measurements at temperatures of 400–540 °C. The molecular structure of the dispersed surface species evolves from isolated monomolybdates (MoO₄ and MoO₅, depending on the support) at low loadings to associated MoO_x units in polymolybdate chains at high loadings and ultimately to bulk crystalline phases for loadings exceeding the monolayer coverage of the supports used. The nature of the oxide support material and of the Mo–O–support bond has a significant influence on the catalytic behaviour of the molybdena catalysts with monolayer coverage. The dependence of reactivity on the support follows the order ZrO₂ > Al₂O₃ > TiO₂ > SiO₂. The oxygen site involved in the anchoring Mo–O–support is of relevance for the catalytic activity.     

Evaluation of the role played by bismuth molybdates in Bi2Sn2O7–MoO3 catalysts used for partial oxidation of isobutene to methacrolein

The catalytic behaviour of multiphasic catalysts based on -bismuth pyrostannate, Bi₂Sn₂O₇, was investigated in the selective oxidation of isobutene into methacrolein. When -Bi₂Sn₂O₇ is mixed with MoO₃, strong cooperation effects on the yield and selectivity in methacrolein occur. However, XRD analyses performed on samples after test revealed the formation of a low quantity of -bismuth molybdate, -Bi₂Mo₃O₁₂, when the reaction temperature exceeded 673 K. Additional experiments were therefore carried out on the “Bi–Sn–Mo–O” catalysts in order to shed light on the role of Bi₂Mo₃O₁₂ in the synergetic effects observed in the Bi₂Sn₂O₇–MoO₃ system. The experimental results are discussed in terms of several hypotheses. First, the intrinsic activity of Bi₂Mo₃ O₁₂ is probably the simplest explanation for the synergetic effects, although experiments have shown that this phase present in a low quantity is only poorly active. Second, catalytic tests made on Bi₂Sn₂O₇–Bi₂Mo₃O₁₂ mechanical mixtures have evidenced a cooperation between these two ternary oxides, particularly when Bi₂Sn₂O₇ was the major component of the mixture. Consequently, it is likely that a synergy between Bi₂Sn₂O₇ and the in situ generated Bi₂Mo₃O₁₂ might play a role in the synergy observed in the Bi₂Sn₂O₇–MoO₃ association. Third, as bismuth pyrostannate was previously shown to behave as an oxygen donor phase with respect to WO₃, a remote control mechanism could therefore occur between Bi₂Sn₂O₇ and MoO₃, independently from the formation of -Bi₂Mo₃O₁₂.     

Maya crude hydrodemetallization and hydrodesulfurization catalysts: An effect of TiO2 incorporation in Al2O3

Hydrotreating of Maya heavy crude oil over high specific surface area CoMo/TiO₂–Al₂O₃ oxide supported catalysts was studied in an integral reactor close to industrial practice. Activity studies were carried out with Maya crude hydrodesulfurization (HDS), hydrodemetallization (HDM), hydrodenitrogenation (HDN), and hydrodeasphaltenization (HDAs) reactions. The effect of support composition, the method of TiO₂ incorporation, and the catalyst deactivation are examined. Supported catalysts are characterized by BET specific surface area (SSA), pore volume (PV), pore size distribution (PSD), and atomic absorption. It has been found that sulfided catalysts showed a wide range of activity variation with TiO₂ incorporation into the alumina, which confirmed that molybdenum sulfided active phases strongly depend on the nature of support. The pore diameter and nature of the active site for HDS, HDM, HDN, and HDAs account for the influence of the large reactant molecules restricted diffusion into the pore, and/or the decrease in the number of active sites due to the MoS₂ phases buried with time-on-stream. The textural properties and hysteresis loop area of supported and spent catalysts indicated that catalysts were deactivated at the pore mouth due to the metal and carbon depositions. The atomic absorption results agreed well regarding the textural properties of spent catalysts. Thus, incorporation of TiO₂ with γ-Al₂O₃ alters the nature of active metal interaction with support, which may facilitate the dispersion of active phases on the support surface. Therefore, the TiO₂ counterpart plays a promoting role to HDS activity due to the favorable morphology of MoS₂ phases and metal support interaction.     

Preparation and structural characterization of Co/Al2O3 catalysts for the ozonation of pyruvic acid

A series of Co/Al₂O₃ catalysts were prepared by the incipient wetness impregnation method using γ-Al₂O₃ support and (CH₃COO)₂Co·4H₂O solutions, followed by calcination at 500–800 °C. Characterization of catalysts was accomplished by several techniques such as thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), physisorption of nitrogen, mercury and helium-based pycnometries, Fourier transform-infrared spectroscopy (FT-IR), X-ray diffraction (XRD) and pH of zero charge (PZC). Impregnation of support produced a moderate decrease of its surface area and pore volume and also led to minor changes of its PZC. Depending on preparation conditions (i.e., calcination atmosphere and temperature and metal loading), one or more of the following Co-containing compounds were identified: CoO, Co₃O₄ and CoAl₂O₄. The support and prepared Co/Al₂O₃ catalysts were tested to catalyze the ozonation of aqueous pyruvic acid at pH 2.5. Pyruvic acid was shown refractory towards single ozonation but the use of γ-Al₂O₃ and Co/Al₂O₃ catalysts resulted in 56–96% pyruvic acid conversion and 41–78% decrease in DOC after 2 h of ozonation of phosphate-buffered solutions. In the absence of the buffer, conversion rate was enhanced likely as a result of pH increase during the course of the process thus giving rise to the indirect way of ozonation through hydroxyl radicals. Acetic acid was found as the main by-product of pyruvic acid ozonation. Depending on the catalyst used, yield of acetic acid varied from 32 to 49%, values noticeably lower that that obtained from the control non-catalytic ozonation experiment (73%). Differences in catalytic activity amongst the various Co/Al₂O₃ catalysts investigated were attributed to the different Co active phases deposited on the γ-Al₂O₃ surface. The following sequence of increasing activity can be inferred from experimental results: CoO, CoAl₂O₄ and Co₃O₄. All the Co/Al₂O₃ catalysts prepared showed good stability as the percentage of cobalt leached out was rather low.