Novel 3D-Printed Alumina Monolith Catalysts for Hydroconversion of Tar

This paper reports on a novel method for the preparation of monolith catalysts using a 3D-printed matrix. The development included an investigation into potential approaches to improving the strength of alumina catalysts, 3D printing of a polymer matrix (template) with a specific channel structure, preparation of templated monolith catalysts, an examination of their physicochemical properties, and testing of these catalysts in hydroconversion of tar. Using an indirect template method, this study is the first to prepare a Al2O3 monolith catalyst with a Schwartz surface microstructure. The extrudate drying rate was found to be the most important parameter for the synthesis of high-strength catalysts. The activity of the monolith catalyst proved to be comparable—and in some parameters even markedly superior—to that of similar granular samples. The tar hydroconversion product consisted of super heavy oil with 2.8 wt % of sulfur.


INTRODUCTION
In Russia oil refining is among the economically most important national industries. Russia annually produces about 300 million tons of petroleum products [1], which represents slightly above a fifth of the total crude oil produced in the country. First of all, this is associated with the oil conversion ratio, and in particular the low level of oil conversion into fuels and lubricants: the average oil conversion ratio at Russian refineries does not exceed 85%, compared to 94-98% in Europe and the USA [2]. On the other hand, given the depletion of the upper (lighter) oil layers, the quality of the crude oil produced has been continuously deteriorating. Therefore, unless effective oil refining technologies are developed, in several decades we will face the inevitable technological dependence on foreign companies, which have already reached conversion ratios of at least 65% even for oil residue feedstocks [3].
To enhance oil refining efficiency, in particular in secondary refining of residues and heavy oils [4], novel types of processes and materials need to be developed. In this context, the best advances are most likely to be achieved in processes that provide the highest yields of hydrocarbon products, namely hydroconversion [5].
Designing a method for the preparation of monolith catalysts that would feature a developed microchannel system (100-800 m in diameter) and polymodality (primarily mesopores and macropores) could be considered a very promising solution. An important advantage of monoliths is the high contact area of the catalyst with the reactor walls, with other monoliths, and with catalytic particles inside the block, thus improving heat transfer inside the monoliths. This eliminates the need to fill chemical reactors with ballasts like silicon carbide to increase the contact surface both between the catalyst particles and between the reactor and the catalyst. Moreover, the well-developed microchannel system would improve the mass transfer of highly NOVEL 3D-PRINTED ALUMINA MONOLITH CATALYSTS viscous feedstock macromolecules. To produce monolith catalysts with a developed system of regularly structured microchannels, this paper proposes an approach based on additive technologies [6]. As the main technique for the preparation of these microchannel catalysts, we used an indirect 3D printing procedure consisting of a series of process steps.
The first step is designing a digital model. This model can be generated by geometric modeling in various computer-aided three-dimensional design (CAD) software packages. The second step, referred to as slicing, involves creating a tessellation pattern by converting a digital model file into a suitable format, most commonly STL. An STL file contains information on each surface of the 3D model in the form of V-shaped sections. The third step generates data on cross sections obtained by layer-by-layer partitioning (slicing) of the STL model, as well as other parameters that instruct the 3D printer to print the object [7]. When the 3D printer configuration is completed, 3D printing can be started. The further preparation steps for templated monolith catalysts are similar to those for granules, except that they include filling a 3D template instead of extruding a plastic mass through a die. Subsequent aging and heat treatment to remove the template are designed to produce an alumina monolith material.
To improve mass and heat transfer during hydroconversion, monolith catalysts with Schwarz and Schoen minimal surface microstructures [8] are the most promising. The diffusion and transport advantages of this catalyst type are described and rationalized, using appropriate calculations and mathematical modeling, by Al-Ketan et al. [9]. These structures feature interconnected channels and connectivity of the solid support. Furthermore, they are distinguished by a surface with nominally zero curvature, generally due to the presence of saddle points and their intersections with a positive curvature radius in one direction and a negative curvature radius equal in absolute value in the other.
Thus, the purpose of this study was to develop a technique for preparing a highly-porous monolith alumina catalyst using 3D printing and to test this catalyst in the hydroconversion of tar.
Synthesis of catalytic materials. To prepare hydroconversion catalysts, we initially needed to synthesize a granular alumina catalyst that would be able to retain its geometrical and textural properties at high pressures and high temperatures. Bearing in mind that additional microchannels produced in alumina by the template method could impair the mechanical strength of the blocks, it was important to identify proper conditions for the preparation of a superior-strength material (compared to conventional granular alumina). For a number of reasons, aluminum hydroxide (namely, bayerite) was used as a support precursor. First, γ-and η-Al 2 O 3 phases that are generated during heat treatment space needed (at 600-800°C) and bayerite (at 250-500°C), respectively, are known to be the most suitable as supports or catalysts from the viewpoint of their physicochemical parameters, especially the textural and acidic properties [10]. The choice of the bayerite from PROMCATALYS as a catalyst precursor was further motivated by an intent to use products from Russian catalyst manufacturers. In this study, a molded composite that contained bayerite and a polymeric template was subjected to aging followed by calcination at an elevated temperature (700-800°C) to ensure a partial transition of the molded precursor into a θ-phase to enhance the strength. Moreover, the high calcination temperature reduces the surface acidity, thus improving the catalyst stability under harsh conditions of tar hydroconversion and rapid coking. Given the goal of preparing a material with improved strength, and knowing that these high temperatures dramatically impair the textural properties of the catalyst, various plasticizers were introduced into the composite to stabilize the meso-and macroporosity of the alumina [11]. For this purpose, 5 mL of HNO 3 and up to 10 mL of organic additives, such as ethylene glycol (EG), diethylene glycol (DEG), glycerol (G), and commercial oil (CO), were added to 365 g of aluminum hydroxide, as the components were agitated in a Z-bladed mixer. For better homogenization, 10 mL of ethanol was also added to the mixture. When adequate homogenization was reached, the mass was extruded through a die into cylindrical granules 3-5 mm long and 3 mm in diameter. All samples were dried and calcined at 700°C to generate a metastable alumina phase. Each granule type was then sorted out and ground to identical geometries for measuring the strength characteristics. The samples were labeled as "mL_additive," where "mL" and "additive" are the volume (in mL) and letter notation (as described above) of the organic additive, e.g., 10_EG. The granular materials were analyzed by X-ray diffraction (XRD) using a Siemens D500 diffractometer in monochromatized CuK α radiation (λ = 1.54 Å), and their crushing strength was measured by an MP-9S instrument. The monolith catalysts were further examined by lowtemperature nitrogen adsorption (using an Autosorb-6B-Kr instrument, Quantachrome Instruments, USA); mercury porosimetry (using a Micromeritics AutoPore IV 9500 porosimeter); and temperature-programmed ammonia desorption (with the ammonia desorption rate being recorded using a calibrated Stanford Research System RGA100 mass spectrometer).
To 3D-print the monolith catalysts, we designed a template with a gyroid lattice (D,P-type surface, and G-type since 2012), which can be approximated by the following trigonometric equation: sin x·cos y + sin y·cos z + sin z·cos x = 0. The catalysts were then printed from polylactic acid (PLA) and polyacrylonitrile butadiene styrene (ABS) using a Wanhao Duplicator D6+ 3D printer. The main printing characteristics are presented in Table 1.
Next, the plastic precursor mass was extruded into the template, dried, graphitized, and calcined at 700°C. To graphitize the template, the composite (polymeric template + paste) was thermostated in a drying cabinet at above 300°C. The catalyst preparation procedure is schematically illustrated in Fig. 1.
The monolith prepared in this manner was tested in tar hydroconversion for 96 h using a 250 cm 3 laboratory setup (Fig. 2) at 400°C, 100 atm, and a feed weight hourly space velocity (WHSV) of 0.25 h -1 . The petroleum products were sampled every 24 h and labeled accordingly. The following physicochemical properties were measured both for the initial tar and products: density (on a VIP-2MR vibration-type liquid density meter), viscosity (on an IKA ROTAVISC viscometer), sulfur content (on a FlashSmart 2000 elemental analyzer, Thermo Scientific, USA), and fractional composition (by simulated distillation as per ASTM 7169).

RESULTS AND DISCUSSION
Close scrutiny of the XRD patterns reveals that the selected Al 2 O 3 precursor is a mixture of predominantly bayerite with a small amount of boehmite (Fig. 3, left). The precursor sample was calcined at 700°C; the XRD pattern is provided in Fig. 3   Developing an effective technique for the preparation of granular catalysts with improved strength is an urgent challenge to be overcome to ensure further advances in 3D printing of monolith catalysts. The mass pushed under pressure into the template needs to be well peptized and molded. To solve this problem, a number of samples were synthesized under identical conditions; these samples contained a total of 10 vol % of various organic additives to increase the plasticity of the precursor mass. The crushing strength of the calcined samples was measured (Fig. 4). The diagram clearly shows that the addition of plasticizers markedly enhances the strength of granular catalysts. Organic additives probably slow down the aging of extrudates, thus extending the granule drying time and improving the binding at microparticle contacts. The best strength-enhancing effect was achieved by adding 5 vol % of diethylene glycol and 5 vol % of ethylene glycol. Therefore, this precursor mixture composition was chosen for further investigation of the monolith synthesis.
The investigation of the granular catalyst demonstrated that drying rate is the critical parameter for the preparation of a high-strength material. Drying of 3×5 mm extrudates is a simple and unchallenging procedure because of their relatively small size. In the case of monolith catalysts, however, an obvious problem is to ensure uniform and controlled drying of the precursor mass introduced into the template if the total volume of this composite exceeds 200 cm 3 . To overcome this challenge, we proposed a special procedure for the drying and calcination of samples (Fig. 5). This procedure consists of a series of steps: drying at room temperature for several days; drying at elevated temperature; graphitization of the template; and calcination of the mass to produce the required Al 2 O 3 phase. Almost all sets of conditions imposed for the drying step failed to achieve the desired result. The prepared monoliths lost their shape even before the end of calcination. The only temperature profile that enabled us to produce high-strength monolith catalysts in the series of experiments was drying at 150°C and graphitization of template at 350°C (Fig. 6).
An investigation of the prepared monolith catalyst by low-temperature nitrogen adsorption, mercury porosimetry, and temperature-programmed ammonia desorption showed the following properties: a specific surface area (S BET ) of 168 m 2 /g, equal mesopore and macropore volumes (V meso = V macro = 0.36 cm 3 /g), and total acidity of 0.188 mmol NH 3 /g. The moderate acidity and the macroporosity satisfy the requirements catalysts must meet for the hydroconversion of heavy oil feedstock such as tar. The scanning electron microscopy (SEM) images (Fig. 7) also indicate the presence of macropores typical of alumina obtained from the above-mentioned precursor.
The monolith Al 2 O 3 sample was subjected to catalytic testing in the first tar hydroconversion step. The main purpose of this step is to reduce the asphaltene content, viscosity, and density of crude oil. The test was carried out over a period of 96 h with sampling every 24 h. The density, viscosity (Table 2), sulfur content, and fractional   (Fig. 8) of the petroleum products were measured. Within the first 24 h, the catalyst exhibited high initial activity in hydrocracking and hydrotreatment. This activity can be explained by the accessibility of the catalyst's active sites, which had not yet been coked. As the catalyst continued running, its activity declined in all processes due to blockage of mesopores by coke deposits. This is typical of hydroconversion catalysts that operate at high temperatures with high-molecular-weight hydrocarbon feedstock [12]. After 72 h, the catalytic activity reached a plateau, and steady-state performance was established. The 3D-printed monolith catalyst with a Schwartz surface microstructure enabled us to decrease the density, viscosity, and S content of the oil tar by 6.7, 99.8, and 17.6%, respectively. After the hydroconversion, the petroleum product contained about 20 wt % of gasoline and diesel fuels, compared to 0.5 wt % in the initial tar feed.
Moreover, the content of vacuum and nonelutable residues dropped by 40%, specifically from 77.1 to 46.3 wt %. This activity is on par with similar known granular catalysts [13,14].

CONCLUSIONS
The composition and heat treatment conditions of extrudates were found to be the most important parameters for the synthesis of high-strength alumina catalysts. The effects of organic additives on the strength of bayeritederived granular Al 2 O 3 catalysts were investigated. The best mechanical properties were exhibited by a material calcined at 700°C with the initial paste containing 5 vol % of ethylene glycol and 5 vol % of diethylene glycol.
The investigation data enabled us to develop a template method for the synthesis of Al 2 O 3 monolith catalysts using a 3D-printed polymer matrix. The preparation   procedure involves the following steps: modeling a template geometry; slicing; 3D printing; extruding a precursor paste into a template; aging; heat treating; and removing the template. It was revealed that, to ensure adequate strength of the monoliths, the composite (paste and polymeric 3D template) must be aged first at room temperature, then at 150°C, followed by graphitizing the template at 350°C. The graphitized template can be completely removed at the required temperature, in this case at 700°C.
The monolith catalyst sample was tested as a guard layer in hydroconversion of tar, a process mainly designed to reduce the viscosity, density, and asphaltene content of the petroleum product. The product consisted of super heavy oil with 2.8 wt % of sulfur when the catalyst reached steady-state performance.
This study is the first to prepare a monolith Al 2 O 3 catalyst with a Schwartz surface microstructure. Also, this material has been tested for the first time in a catalytic process, namely in tar hydroconversion. The activity of the monolith catalyst in the hydroprocessing of oil residues was comparable to that of granular alumina with a similar composition.