Search

Hydroisomerization and hydrocracking of octylcyclohexane (C14H28), obtained after in situhydrogenation of the parent aromatic compound, phenyloctane (C14H22), on a Pt/Al2O3precatalyst, were performed over a bifunctional Pt/USY zeolite catalyst, at 573 K, 6 MPa, and a molar ratio H2/HC of 7 mol/mol. About 200 reaction products were identified and quantified with GCxGC–FID/MS and then lumped into families according to carbon number and chemical similarity. Insight into reaction pathways was gained from the evolution of the composition of the isomers and cracked products with conversion. The results illustrate a kinetic competition between isomerization steps, which increase the number of branchings on the ring by shifting carbon atoms from the long alkyl chain to the ring, with exo-cyclic cracking of the long alkyl chain. Favorable cracking distributions are reached as soon as the molecule is tribranched, which leads to a peak at C7 in the product distribution.

The aim of this study was to understand the impact of vacuum gas oil (VGO) properties on the deactivation rate of a hydrocracking catalyst (nickel–molybdenum sulfide dispersed on a carrier containing USY zeolite). For this purpose, two hydrotreated feeds of different densities, organic nitrogen (∼120–150 ppmw) and aromatic content, were hydrocracked under operating conditions that favor catalyst deactivation, that is, high temperature (T= 418 °C) and high space velocity (LHSV = 3 h–1). The catalyst performance was followed by measuring the VGO conversion (370 °C+petroleum cut) and determining the apparent kinetic constants for the main hydrocracking reactions (cracking, hydrodenitrogenation, hydrodesulfurization, and aromatics hydrogenation). The experiments were stopped after different times on stream (either 6 or 30 days) in order to assess the evolution of the catalyst as a function of time. The spent catalysts, obtained from three different reactor locations, were characterized by elemental and textural analyses and by thermogravimetry to investigate the quantity and nature of the coke formed. Catalytic tests with different model compounds (toluene and n-heptane) were carried out to determine the residual activity of the hydrogenating and acid catalyst functions. It was found that, at the evaluated conditions, both the nature and the content of organic nitrogen and aromatics compounds of the feedstock have a determinant role in the deactivation rate. Organic nitrogen determines the ratio between available metal and acid sites. The aromatics generate coke precursors on the available acid sites. Both factors play a coupled role that promotes coke deposition on the catalyst surface, which leads to an increase in the deactivation rate on top of the end boiling point of the feed.

An innovative approach using solid/liquid interface characterization (zetametry measurements, ATR-IR), coupled with ex situFTIR has been used to build a phosphate complexation model at the alumina surface via the PHREEQC software. According to the phosphorus coverage impregnated on the alumina surface, two mechanisms of adsorption have been identified: a specific surface complex adsorbed on the predominant alumina facet and a surface precipitation process. At low phosphorus coverage (P coverage < 1.2 at/nm2), the adsorption proceeds mainly through a monodentate complex adsorbed on μ1-AlIV-OH sites located on the (110) facet, and a bidentate involving μ1-AlV-OH (110) is also discussed. However, the surface concentration of these particular sites is limited. Thus, above the threshold of 1.2 at/nm2, a surface precipitation occurs with the formation of an amorphous AlPO4phase, experimentally observed by FTIR and zetametry measurements.

The adsorption processes for postcombustion CO2capture are usually based on a temperature or vacuum swing (TSA or VSA). In the present contribution an alternative concept is presented, which is based on the regeneration of the solid sorbent (an immobilized amine) by a purge gas, low pressure vapor, under almost isothermal conditions. Because of the close to isothermal operation, the process consumes significantly less thermal and mechanical energy than conventional TSA and VSA processes, respectively. We present a rough design of such an isothermal concentration swing process based on a simplified, analytical model of a 2-step PSA (pressure swing adsorption) process. The analytical model allows a definition of the range of operating conditions that lead to the best compromise between energy consumption and productivity (size and number of the adsorbers). Moreover, it is possible to define the equilibrium and mass transfer properties of the ideal solid sorbent. The feasibility of the concentration swing process was finally validated by numerical simulations of a full PSA cycle under adiabatic conditions.

Hydroconversion of perhydrophenanthrene was performed over Pt/Beta and Pt/ASA bifunctional catalysts and compared to results obtained on the Pt/USY zeolite catalyst, under the same operating conditions. Perhydrophenanthrene resulted from the hydrogenation of the parent aromatic, phenanthrene, on a Pt/alumina pre-catalyst. All three bifunctional catalysts followed the general reaction pathway observed previously, i.e.isomerization of perhydrophenanthrene by ring-shift and ring-contraction, followed either by the formation of alkyladamantanes or by ring-opening. The ring opening products cracked to smaller naphthenes. Yet, the intermediates of this general reaction network differed very clearly from one catalyst to another. These shape selectivity effects could be explained by GCMC simulations of the adsorption selectivities of different intermediates. Bulky intermediates were preferentially adsorbed on USY zeolites, whereas Beta zeolites adsorbed preferentially linear structures. The product distribution on Beta zeolites was explained through the formation of a central ring-opening intermediate which cracked rapidly to C7naphthenes. USY-based catalysts were the most active among the solids tested, while Beta zeolites were slightly more selective to generate cracked products.

Most industrial hydrocracking catalysts contain metal sulfides. It is commonly accepted that sulfide-based bifunctional hydrocracking catalysts are less well balanced than Pt-based catalysts because of their weak (de)hydrogenation activity. However, the difference in catalytic performances between the two catalysts has never been quantified. In the present work, hydrocracking catalysts were prepared by shaping the USY zeolite CBV720 with an alumina binder. Acidity was varied by changing the USY content. The extrudates were impregnated either with platinum or with nickel-molybdenum-phosphate and evaluated in the hydroconversion of n-hexadecane. In the absence of NH3, NiMo sulfide was unable to supply the acid sites and to avoid cracking reactions. Cracking became the dominating pathway. In the presence of NH3, only a tiny fraction of the acid sites was vacant. This allowed the NiMo sulfide catalysts to become fairly well balanced, if the zeolite loading was low. Increasing zeolite loading led to a deviation from ideality. Pt catalysts were, with one single exception (at very high USY content), always balanced, even in the absence of NH3.

This paper aims to understand the reactivity of different nitrogen species during the catalytic hydrodenitrogenation (HDN) of a mixture of straight-run gas oil and coker gas oil. Effluents at different HDN conversions were analyzed by Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR/MS), with electrospray ionization (ESI). Positive ESI (ESI(+)) allows a selective ionization of basic nitrogen compounds, whereas negative ESI (ESI(−)) leads to selective ionization of neutral nitrogen compounds. FT-ICR/MS in ESI(+) or (−) mode generates distributions of the nitrogen compounds in the basic and neutral class, respectively, according to their carbon number and double bond equivalent (DBE), which is related to the degree of aromaticity and the number of rings. The evolution of the DBE distribution and the carbon number distribution as a function of HDN conversion gives rich information concerning the main reaction pathways and the reactivity of different nitrogen species. For the basic compounds, a shift to a lower DBE was observed, which was interpreted as the formation of partially hydrogenated or ring-opened intermediates. These intermediates were then slowly converted to the final HDN products. At intermediate conversion levels, especially light compounds were accumulated as intermediates whereas the heavy compounds were directly converted to HDN products, probably due to preferential adsorption. The neutral compounds showed a very different behavior. At the early reaction stages, they were quickly converted to HDN products, but at high conversion, the conversion of residual carbazole and tetrahydrobenzocarbazole compounds was completely inhibited. The inhibition was probably provoked by the formation of partially hydrogenated basic intermediates, which were stronger inhibitors than the aromatic pyridine rings in the feed. The contribution of cracking reactions was weak, since the overall carbon number distribution of the nitrogen compounds did not change much during hydrotreatment.

Two NiMo sulfide catalysts, supported on Al2O3and silica–alumina, respectively, were compared in the hydrotreatment of a mixture of straight-run and coker gas oil. Experiments were conducted in a continuous fixed bed reactor, at 623–643 K and 5 MPa. The Al2O3-supported catalyst proved to be slightly more active in HDN and significantly more active in HDS than its silica–alumina analogue. The activity trend was attributed to the inhibition provoked by the strong adsorption of basic nitrogen species. Earlier work had shown that the adsorption (and thus the inhibition) was significantly stronger on an acidic support. Our data further showed that the reactivity of neutral (pyrrole benzologues) and basic nitrogen compounds was different. The basic fraction was converted in a gradual fashion, whereas the large majority of neutral species was converted very quickly, except for a small fraction, which remained practically inert to hydrotreatment.

Pressure swing adsorption (PSA) is a promising method for an economically viable separation of CO2from gas mixtures. One important step in developing efficient PSA separations is to choose an adsorbent with a good compromise between selectivity and working capacity. The problem can be seen as an optimization process where one searches the properties of the adsorbent that give the best trade-off between selectivity and working capacity. In a recent work we have developed a method to optimize the properties of nonpolar adsorbents for PSA separation of CO2. In this work we have extended the analysis to polar adsorbents. A simplified model of the electric field generated by the solid is used to gain a better understanding of the effect of the geometry and the magnitude of the field on the adsorption behavior of the system. General rules about optimal adsorbents for CO2adsorption in porous solid can be drawn. These rules allow defining a strategy for the intelligent design of porous materials for CO2separation applications.

In this work, we propose to evaluate and validate an emerging spectroscopic space-resolved technique: atomic force microscopy coupled with infrared spectroscopy (AFM–IR) for inorganic materials in tapping mode at the nanoscale. For this aim, a preliminary investigation of sample preparation techniques was done and the stability of tapping AFM–IR spectra was evaluated on reference samples [poly(methyl methacrylate) and silica]. It was concluded that for a homogeneous polymer, it is possible to compare AFM–IR spectra with conventional Fourier-transform infrared (FTIR) spectra obtained in transmission. When an inorganic solid is considered, AFM–IR spectra are different from the global FTIR spectrum which indicates that the AFM–IR technique probes a volume which is not representative of global composition, that is, the external surface layer. Moreover, local infrared spectra recorded in the tapping mode of the external surface are significantly different depending on the analyzed regions of the same particle and between particles of the amorphous silica, implying surface heterogeneity. The AFM–IR technique allows surface description of amorphous inorganic materials at the nanoscale and opens new frontiers in the characterization of functional nanoscale and crystalline materials.

The catalytic performance of a hydrocracking bifunctional catalyst and consequently its activity loss depend on the balance between the metal (M) and acid (A) functions. Commonly, the metal function is reputed to stabilize catalyst activity and should therefore alleviate deactivation. In order to test this hypothesis, this study aimed to assess the impact of metal oxide (MoO3) content on the deactivation of a vacuum gasoil hydrocracking catalyst (nickel-molybdenum sulfide dispersed on a carrier containing USY zeolite). To do so, two bifunctional catalysts with different metal oxide contents were submitted under specific operating conditions promoting catalyst deactivation, i.e., high temperature (T= 418 °C) and high space velocity (LHSV = 3 h–1). Experiments were carried out in a pilot unit with a hydrotreated VGO feedstock containing 120 ppmw of organic nitrogen. The loss in activity was measured by following the decrease of both the VGO conversion (370 °C+) and the apparent kinetic constants for the main hydrocracking reactions (cracking, hydrodenitrogenation, and aromatics hydrogenation) with time on stream. The aged catalysts, sampled from three different reactor locations, were characterized by elemental and textural analyses and thermogravimetry to investigate the quantity and nature of the coke deposits. To determine the residual activity of the hydrogenation and acid catalyst functions, catalytic tests with different model compounds (toluene and n-heptane) were carried out. It was observed that in the hydrocracking of feedstocks with high nitrogen concentration, the decrease of the metal content causes a substantial increase of the nitrogen compounds’ concentration in the catalytic bed. This increase significantly reduces the available acid sites and hence the hydrocracking activity. However, at the same time, the M/A ratio is modified in favor of the metal function. The better balance of the catalyst with low metal loading results in a lower coke deposition and consequently a lower deactivation rate than with the catalyst with higher metal content.

A systematic study of the effect of coordinatively unsaturated sites (cus) in the separation of CO2/CH4and CO/CO2/CH4mixtures on CPO-27-M (M = Ni, Co, and Zn) and STA-12-Ni metal–organic frameworks was carried out using gravimetric and breakthrough experiments. The separation selectivity and the working capacity of these structures were evaluated as important performance indicators for CO2separations by PSA. The results demonstrate a remarkable influence of coordinatively unsaturated sites on the selectivity and the working capacity. Particularly, the high affinity of CPO-27-Ni and CPO-27-Co for CO2leads to a low working capacity for CO2(because regeneration of the adsorbents is difficult) but a high CO2/CH4selectivity. With a ternary CO/CO2/CH4feed mixture, CPO-27-Ni and -Co prefer the adsorption of CO over CO2due to the strong specific interaction of CO with cus. Surprisingly, STA-12-Ni does not exhibit the same behavior: it is selective for the adsorption of CO2in a ternary mixture, just like CPO-27-Zn. Among the four MOFs tested in this study, CPO-27-Zn presents the best compromise between the working capacity and the CO2/CH4and CO2/CO selectivities. The results are discussed in terms of the coordination chemistry of the coordinatively unsaturated metal sites, their acid–base properties, and their accessibility.

The separation of xylene isomers is one of the most difficult separations in petrochemistry. Adsorbents have to preferentially adsorb para or meta isomers, but there are only a few zeolites which fulfill the criteria of selectivity and adsorption capacity. In this study, we evaluate two metal–organic frameworks (MOF) with coordinatively unsaturated metal sites (cus), i.e., CPO-27-Ni and HKUST-1, in the adsorption and separation of p-xylene, m-xylene, and o-xylene in the gas phase. The results are compared with those of the zeolite NaY. CPO-27-Ni and HKUST-1 are both ortho-selective adsorbents, but molecular simulations indicate that the reasons for the ortho selectivity are quite different. The CPO-27-Ni structure is intrinsically ortho selective, and the ortho selectivity is further enhanced by electrostatic effects. HKUST-1, on the other hand, is intrinsically (weakly) para selective but the para selectivity is overcompensated by electrostatic effects, which overall leads to a slight preference for the ortho isomer adsorption. The insights in the adsorption behavior of xylenes that are provided by this study provide useful guidelines in the quest for para-selective MOFs materials.

This work reports the adsorption and coadsorption data of CO2/CH4/CO mixtures on several metal–organic frameworks [MOFs; MIL‐100(Cr), MIL‐47(V), MIL‐140(Zr)‐A, Cu‐btc, and MIL‐53(Cr)] and compares them with reference adsorbents, that is, zeolite NaX and an activated carbon material, AC35. We also evaluate the effect of H2O on CO2adsorption and on the stability of the structures. Based on the experimental adsorption data, the performance potential of MOFs in several pressure swing adsorption processes is estimated by making a ranking of working capacities and separation factors. We discuss the separation of biogas, the purification of H2produced by steam reforming of methane, and the removal of CO2from synthesis gas in IGCC (integrated gasification combined cycle) systems. Some MOFs are very well placed in the ranking of (isothermal) working capacity vs. selectivity. Yet, performance is not the only criterion for the selection of MOFs. Ease and cost of synthesis and long‐term stability are other important aspects that have to be taken into account.

The formation of the iron species in Fe/ZSM-5 prepared by chemical vapour deposition (CVD) of FeCl₃ onto H/ZSM-5 was monitored by characterising the catalyst after each step of its preparation by X-ray absorption, UV–VIS and atomic absorption spectroscopy (AAS), and by magnetic susceptibility measurements. Upon the CVD treatment, the iron ions are exchanged in the zeolite pores in the form of tetrahedrally coordinated [FeCl₂]+ species. During the subsequent washing of the zeolite most of these ions (ca. 80%) aggregate into small binuclear hydroxo-iron clusters with a structure similar to that of the binuclear units in α-FeOOH. The remaining ions are present as isolated species, while a minor fraction of multinuclear clusters may be present as well. Calcination in oxygen after the washing results mainly in the removal of the water molecules coordinatively bound to the iron species. Both the washing of the zeolite and its calcination in oxygen were found to be of fundamental importance for the formation of small clusters. When the washing is omitted, the calcination yields relatively large Fe₂O₃ particles. If the thermal treatment after the washing is carried out in helium rather than in oxygen, partial autoreduction of the iron from Fe3+ to Fe2+ takes place together with the formation of Fe₃O₄ clusters. The possible structure of the binuclear (hydr)oxo-iron clusters is discussed.