Development of a Novel Structured Mesh-Type Pd/γ-Al₂O₃/Al Catalyst on Nitrobenzene Liquid-Phase Catalytic Hydrogenation Reactions

Abstract

Nitrobenzene liquid-phase catalytic hydrogenation is commonly regarded as one of the most effective technologies for aniline production. The traditional granular catalysts have the disadvantages that the reactor bed pressure drop is large and the mass transfer efficiency between gas and liquid phases is low. In this study, a novel structured mesh-type Pd/γ-Al₂O₃/Al catalyst was prepared by anodic oxidation and pore structures of γ-Al₂O₃/Al supports were constructed by acid pore-widening treatments. The results showed that acid pore-widening treatments can improve the pore size of γ-Al₂O₃/Al supports; the support with HNO₃ pore-widening treatment exhibited the largest pore size, being enlarged from 3.7 nm to 4.6 nm. The Pd/γ-Al₂O₃/Al catalysts prepared with different acid pore-widening treatment supports contribute to the increased active metal Pd loading, more Pd⁰ content, and better dispersion of the Pd particles. The catalyst prepared with HNO₃ pore-widening treatment support exhibited the largest active metal Pd loading, enlarging from 1.82% to 1.95%, the largest Pd⁰ content being enlarged from 52.1% to 58.5% and the smallest Pd particle size being reduced from 103 nm to 41 nm, resulting in the highest nitrobenzene conversion, increasing from 67.2% to 74.3%. Eventually, we calculated that the pressure drop of structured catalysts was 1/72 of that of granular catalysts, resulting in a better diffusion of the H₂ through nitrobenzene solution to active sites on the catalyst surface and a significant increase in the catalytic activity.

1. Introduction

Aniline, as a vital raw material and key intermediate for many chemical products [1], is an important organic compound with a wide range of applications in the chemical industry, pharmaceuticals, and dyestuffs [2]. Furthermore, aniline serves as a raw material for the production of polyurethane raw material diphenylmethane diisocyanate (MDI), having a great market potential [3,4]. At present, the production methods of aniline mainly consist of nitrobenzene iron powder reduction [5], phenol amination [6], and nitrobenzene catalytic hydrogenation [7]. Among them, nitrobenzene catalytic hydrogenation has advantages of low pollution and high product quality, accounting for 85% of the total production capacity of aniline [8]. The chemical reaction of nitrobenzene catalytic hydrogenation reduction to aniline is [9]:

$${\mathrm{C}}_6{\mathrm{H}}_5\mathrm{N}{\mathrm{O}}_2+3{\mathrm{H}}_2\to {\mathrm{C}}_6{\mathrm{H}}_5\mathrm{N}{\mathrm{H}}_2+2{\mathrm{H}}_2\mathrm{O}$$

(1)

The procedure of nitrobenzene catalytic hydrogenation mainly consists of gas-phase catalytic hydrogenation [10] and liquid-phase catalytic hydrogenation [11]. Gas-phase catalytic hydrogenation is widely used due to its high conversion, whereas it is greatly hindered by the stringent operating conditions (high temperature and pressure) and unavoidable by-products (nitroso and azo compounds) [12]. On the other hand, liquid-phase catalytic hydrogenation exerts better catalytic effects, including fewer by-products, high selectivity, low hazard factor, easy separation of products, and low energy consumption [13,14], and has gradually become the mainstream technology for aniline production.

In liquid-phase catalytic hydrogenation, the mixing and mass transfer between the gas and liquid phases is one of the most important factors affecting the hydrogenation efficiency and product conversion [15,16,17]. At present, most of the catalysts used for nitrobenzene liquid-phase catalytic hydrogenation reaction are granular catalysts, in which the reactor bed pressure drop is large and the mass transfer efficiency between the gas and liquid phases is low, leading to limitations in the improvement of aniline production. In recent years, the application of structured catalysts in multiphase catalysis has attracted the attention of numerous scientists [18]. Structured catalysts can significantly improve the mass and heat transfer performance of the catalyst bed, optimize the structure of the flow field, reduce the pressure drop, and lower the reaction hazard factor [19]. To be more specific, 3D-printed catalytic static mixers (csm) [20] had great achievements in various hydrogenation reactions [21,22]. In addition, researchers [23,24,25] also prepared aluminum oxide powder coating catalysts to reduce pressure drop and strengthen stability. However, most of the 3D-printed catalytic static mixers are limited in size and are not suitable for mass production applications [26]. Meanwhile, aluminum oxide powder coating can easily fall off of the aluminum oxide powder coating catalyst substrate, especially in reactions including liquid phases [27]. Therefore, such dilemmas require us to develop a catalyst with stronger bonding to the substrate and greater production capacity.

In this work, structured mesh-type γ-Al₂O₃/Al supports were prepared by anodic oxidation technology and then structures of various pore sizes were adjusted by acid pore-widening treatments. After loading, the Pd/γ-Al₂O₃/Al catalysts were applied to nitrobenzene liquid-phase catalytic hydrogenation reaction in a fixed-bed reactor. Firstly, the various pore physical properties of structured mesh-type γ-Al₂O₃/Al supports were investigated by SEM and BET, and the differences in pore structure of a series of acid pore-widening supports were further explored. Secondly, the effects of acid pore-widening treatments on the distribution of metal Pd particles and Pd⁰ content of the catalysts were further investigated by SEM, BET, XPS, and XRD. Thirdly, the activities of granular catalyst and structured mesh-type catalyst as well as different acid pore-widening treatment catalysts in nitrobenzene liquid-phase catalytic hydrogenation reactions were investigated. Eventually, to investigate the mass transfer-enhancing performance of the structured mesh-type catalyst, the pressure drop difference between the granular catalyst and structured mesh-type catalyst was compared through calculation in a nitrobenzene liquid-phase catalytic hydrogenation reaction system.

2. Materials and Methods

2.1. Catalyst Preparation

The structured γ-Al₂O₃/Al support was prepared by anodizing technology. Firstly, the commercial Al mesh (22 cm × 28 cm) was soaked in 10 wt% Al mesh cleaner for 7 min to remove oil stains on the surface of the Al mesh, then washed with deionized water and air-dried. Secondly, the aluminum mesh was anodized in 104 L of a 0.4 M oxalic acid solution under a current density of 25 A/m² and a constant temperature of 16 °C for 10 h in order to obtain the anodic aluminum oxide (AAO). The AAO was calcined at 350 °C in a muffle furnace to decompose residual oxalic acid on the surface. Thirdly, the AAO was hydrated in 80 °C deionized water for 1 h. Finally, the hydrated AAO was dried at room temperature followed by calcining at 500 °C in a muffle furnace for 4 h to form γ-Al₂O₃/Al support.

The active components of Pd were loaded by impregnation. Firstly, the as-prepared γ-Al₂O₃/Al was impregnated in 400 mL of 1.3 g/L H₂PdCl₄ solution for 24 h, and then dried and calcined at 550 °C for 4 h. Secondly, the calcined catalyst was reduced in 0.05 mol/L sodium borohydride solution for 4 h.

2.2. Acid Pore-Widening Treatment on Support

The pore structures of the γ-Al₂O₃/Al supports were constructed by acid pore-widening treatments. Firstly, the prepared γ-Al₂O₃/Al supports were impregnated in 10 wt% nitric acid, sulfuric acid, or oxalic acid solutions for 30 min. Secondly, the supports with different acid pore-widening treatments were calcined at 350 °C to remove the different acids on the surface.

The active components of Pd were also loaded by impregnation. The γ-Al₂O₃/Al supports with different acid pore-widening treatments were impregnated in 400 mL of 1.3 g/L H₂PdCl₄ solutions for 24 h, and then dried and calcined at 550 °C for 4 h. Then, the calcined catalysts were reduced in 0.05 mol/L sodium borohydride solution for 4 h.

2.3. Catalyst Characterization

The actual contents of Pd on the catalyst were measured by inductively coupled plasma atomic emission spectrometry (Agilent, Santa Clara, CA, USA). Nitrogen adsorption was performed on an ASAP 2020-M instrument (Micromeritics, Norcross, GA, USA). The Brunauer–Emmett–Teller (BET) surface area (S_BET) of the catalyst was obtained by the nitrogen adsorption method after degassing at −196 °C for 6 h, then the pore volume (Vp) and pore size distribution (Dp) were determined by the Barret–Joyner–Halender (BJH) method using desorption data. Scanning electron microscopy (Hitachi, Tokyo, Japan) was performed to characterize the catalyst surface morphology. To improve the sharpness of the imaging, the platinum plating treatment was carried out in advance. X-ray polycrystalline diffractometry (Bruker, Billerica, MA, USA) was performed using a Cu Kα source to obtain the phases of crystals on the catalysts. X-ray photoelectron spectroscopy (Thermo Fisher Scientific, Waltham, MA, USA) was performed to analyze the surface chemistry of the catalyst, and the sample was corrected for the charge energy shift by the C 1s peak.

2.4. Catalyst Activity Test

The flow chart of the nitrobenzene liquid-phase catalytic hydrogenation experimental device is shown in Figure 1. Nitrobenzene solution and hydrogen were used as raw materials for the reaction and entered the reaction tube through the control of a pump and gas flow meter, respectively. The concentration of the nitrobenzene solution was 0.03 M and the solvent was 70% concentration of aqueous ethanol. The flow rate ratio of hydrogen to nitrobenzene solution was 10:1. Granular and structured mesh-type catalysts were, respectively, filled into a reaction tube (12 mm diameter, 100 mm length) within the fixed bed to form a catalyst-filled gas–liquid–solid fixed-bed reactor. The granular catalyst was made by shearing the mesh-type catalyst into 20–40 mesh particles and mixing it with quartz sand of the same mesh. It is worth noting that the structured catalyst was a mesh-type shaped small circular plate with a diameter of 12 mm and a thickness of 0.4 mm. The diameter of the mesh-type catalyst was sized to fit the reaction. A total of 1.3 g of catalyst was used for each test (24 pieces of catalysts in total for mesh-type catalyst). All catalysts did not require activation. The conditions for activity evaluation were normal temperature and pressure. The reaction products were detected and analyzed by gas chromatography (GC-2014, Shimadzu, Japan). The chromatographic column model was SH-I-624Sil.

The conversion of nitrobenzene and the yield of aniline were calculated as follows:

$$\mathrm{Nitrobenzene}\mathrm{conversion}\left(\mathrm{\%}\right)=\frac{C_0-C_1}{C_0}\times 100\mathrm{\%}$$

(2)

$$\mathrm{Aniline}\mathrm{yield}(\%)=\mathrm{Nitrobenzene}\mathrm{conversion}(\%)\times \mathrm{F}\times 0.03\mathrm{M}$$

(3)

where $C_0$ and $C_1$ are, respectively, the concentration of nitrobenzene solution at the inlet and outlet of the reactor. $\mathrm{F}$ is the flow rate of nitrobenzene solution.

3. Results

Figure 2a shows the general appearance of the commercial Al mesh. Figure 2b shows the cross-sectional morphology of AAO which was observed by optical microscope. It can be seen that a porous and fluffy alumina film (the white section) with a thickness of 86.2 μm grew vertically at the surface of the Al substrate (the blue section) after anodizing. Additionally, a 10 μm dense layer between the Al layer and the alumina layer ensured the tight attachment of the catalyst layer to the substrate compared with the traditional aluminum oxide powder coating, which is more promising for application in gas–liquid reaction systems. Figure 2c,d shows the SEM images of AAO and γ-Al₂O₃ surfaces and the surface of AAO composed of a series of ordered pore structures was observed. Compared to AAO, the surface of γ-Al₂O₃ after hot water combining treatment and calcination became rougher and the distribution of pores become disordered, consistent with the previous study [28].

Figure 3a–d show the surface of γ-Al₂O₃/Al supports with different acid pore-widening treatments. Notably, it was found that with short acid impregnation time, the rough surface and disordered distributed pores of the supports were not changed. Moreover, Figure 3e–h show the cross-section of different supports. Compared with untreated γ-Al₂O₃/Al, the pore diameters of the cross-section in acid pore-widening treatment supports became larger, indicating that acid pore-widening treatment had a corrosive effect on the supports. Figure 3i–l show the diagram of the pore diagram distribution. The diagram of pore diameter distribution of different supports after the H₂C₂O₄ and H₂SO₄ pore-widening treatment is similar to the untreated γ-Al₂O₃/Al, but the pore diameter distribution lines are slightly skewed to the right. Compared with the images of (H₂SO₄)γ-Al₂O₃/Al and (H₂C₂O₄)γ-Al₂O₃/Al supports, there are more larger pore diameter distributions of support after the HNO₃ pore-widening treatment, exhibiting the most uniform surface and largest section of pores and indicating that HNO₃ had the most effective pore-designing effect on γ-Al₂O₃/Al support.

The BET analysis was carried out to clarify the pore physical properties of the supports with different acid pore-widening treatments. Figure 4 shows the N₂ adsorption–desorption isotherms of different γ-Al₂O₃/Al supports. Firstly, the Type IV isotherms of both γ-Al₂O₃/Al and acid pore-widening treatment γ-Al₂O₃/Al supports had a hysteresis loop, indicating that the acid pore-widening treatments did not change their mesoporous nature. Moreover, the relative pressure of (HNO₃)γ-Al₂O₃/Al hysteresis loop formation is higher compared to other supports, attributed to the fact that the (HNO₃)γ-Al₂O₃/Al has the largest pore diameter.

Table 1. Textural properties of different acid pore-widening treatment γ-Al₂O₃/Al supports.

Support	SBET (m2/g)	Vp (μL/g)	Dp (nm)
γ-Al2O3/Al	70	97.1	3.7
(H2C2O4)γ-Al2O3/Al	72	99.9	3.8
(H2SO4)γ-Al2O3/Al	75	105.1	4.0
(HNO3)γ-Al2O3/Al	80	116.1	4.6

summarizes the textural properties of the different supports. Generally, the specific surface area and average pore size of γ-Al₂O₃/Al supports increased after all kinds of acid pore-widening treatments. The pore diameters are uncertain below around 5 nm, where the BET and BJH methods are inaccurate. However, it is obvious that the (HNO₃)γ-Al₂O₃/Al showed the largest specific surface area and had the largest pore size of about 4.6 nm, consistent with the results of SEM images and N₂ adsorption–desorption isotherm analysis above.

In this study, Pd/γ-Al₂O₃/Al catalyst was produced by active metal Pd impregnation. Figure 5a–d show that the surface of Pd/γ-Al₂O₃/Al was obviously covered with Pd particle. These small Pd metal particles were uniformly dispersed on the surface of the catalyst, and the catalyst surface exhibited a dense and ordered porous structure. The Pd particles on the surface of the Pd/γ-Al₂O₃/Al without acid pore-widening treatment had larger particle diameter and agglomeration, attributed to the small pore size of the support without acid pore-widening treatment, which was not favorable for the diffusion and attachment of metal particles.

Figure 6a–d show the metal particle size distribution of Pd/γ-Al₂O₃/Al catalysts after different acid pore expansion modifications. The results showed a significant reduction in the particle size of metal particles on the surface of catalysts with different acid pore-widening treatments compared with the untreated Pd/γ-Al₂O₃/Al, indicating that with the increasing pore size of the acid pore-widening treated supports, the active metal particles could be more easily dispersed and adhered during the loading process [29]. Remarkably, given the largest pore size of the (HNO₃)γ-Al₂O₃/Al support compared to (H₂SO₄)γ-Al₂O₃/Al and (H₂C₂O₄)γ-Al₂O₃/Al supports, the surface of Pd/(HNO₃)γ-Al₂O₃/Al support exhibited the smallest average Pd metal particle size, namely 41 nm. On the contrary, owing to the smallest pore diameter increase in the (H₂C₂O₄)γ-Al₂O₃/Al compared to the (H₂NO₃)γ-Al₂O₃/Al and (H₂SO₄)γ-Al₂O₃/Al, the surface of Pd/(H₂C₂O₄)γ-Al₂O₃/Al still showed metal particle agglomeration on the catalyst surface.

The BET analysis was carried out to clarify the pore physical properties of the catalysts after active metal Pd loading. Figure 7 shows the N₂ adsorption–desorption isotherms of different Pd/γ-Al₂O₃/Al catalysts. Firstly, the Type IV isotherms of acid pore-widening treatment γ-Al₂O₃/Al catalysts had a hysteresis loop, indicating that the active metal Pd loading did not change their mesoporous nature. Moreover, the hysteresis loop type of all Pd/γ-Al₂O₃/Al catalysts changed from H1 to H3, indicating that the pore structure of all Pd/γ-Al₂O₃/Al catalysts became significantly irregular after the active metal Pd loading.

Table 2. Chemical composition and structural properties of different Pd/γ-Al₂O₃/Al catalysts.

Catalyst	Pd Loading (wt%)	SBET (m2/g)	Vp (μL/g)	Dp (nm)
Pd/γ-Al2O3/Al	1.82	56	91.4	6.6
Pd/(H2C2O4)γ-Al2O3/Al	1.88	59	93.2	7.4
Pd/(H2SO4)γ-Al2O3/Al	1.90	63	98.5	7.5
Pd/(HNO3)γ-Al2O3/Al	1.95	69	108.6	7.8

summarizes the chemical composition and textural properties of the different catalysts. The Pd loading of acid pore-widening treatment Pd/γ-Al₂O₃/Al catalysts increased compared to the untreated Pd/γ-Al₂O₃/Al, owing to the increase in the specific surface area of the acid pore-widening treatment Pd/γ-Al₂O₃/Al supports. Strikingly, the Pd/(HNO₃)γ-Al₂O₃/Al had the largest Pd loading, owing to the largest specific surface area of (HNO₃)γ-Al₂O₃/Al. The pore diameters of different catalysts increased after loading compared to the untreated Pd/γ-Al₂O₃/Al catalyst, which could be attributed to the fact that the acidic impregnating solution, which produces a certain pore-expanding effect, corrodes the pores during impregnation. The main reason for the pore diameter becoming bigger when the size properties became smaller is that the small pores were blocked (1–2 nm) and the number of pores decreased [29,30,31,32].

Based on XRD, we further investigated the crystalline phases of the catalysts with different acid pore-widening treatments. Figure 8 demonstrates that the XRD diagrams of the catalysts prepared with different acid pore-widening treatments all exhibited three characteristic diffraction peaks, representing the (111) crystal planes of γ-Al₂O₃ and metal Pd⁰ [32]. Those traces also indicate (large linewidths and background structure) that much of the material is sub-crystalline transition alumina. These results also verify the fact that the acid pore-widening treatment did not disrupt the γ-Al₂O₃ structure of the supports as well as the Pd(111) crystal surface of the active metal Pd particles in the catalysts.

It is well-known that charge properties of catalysts are significant in the performance of catalysts [31]. The surface chemistry properties of Pd on catalysts on different Pd/γ-Al₂O₃/Al catalysts were identified by XPS. Figure 9 shows the Pd 3d XPS spectra of different catalysts. The XPS spectra of Pd 3d can be divided into Pd⁰ peak (orange section) and Pd²⁺ peak (green section). The sample was corrected for the charge energy shift by the C 1s peak (284.8 eV) [33]. The binding energies at 335 eV and 340.3 eV were, respectively, assigned to Pd 3d_5/2 and Pd 3d_3/2 of Pd⁰ particles in the catalysts prepared with different acid pore-widening treatments, indicating the presence of metallic Pd⁰ loaded onto the γ-Al₂O₃/Al supports. The second pair of Pd signals appearing at 336.3 eV and 341.9 eV was attributed to Pd atoms with lower charge density (Pd²⁺) [34].

To visually observe the amount of Pd in different valence states on different catalysts, we conducted further investigation.

Table 3. Chemical compositions of different Pd/γ-Al₂O₃/Al catalyst surfaces obtained by XPS calculations.

Catalyst	Pd (%)
	Pd0	Pd2+
Pd/γ-Al2O3/Al	52	48
Pd/(H2C2O4)γ-Al2O3/Al	54	46
Pd/(H2SO4)γ-Al2O3/Al	56	44
Pd/(HNO3)γ-Al2O3/Al	59	41

summarizes the relative contents of Pd in different valence states on different catalysts, which were calculated from the curve fitting in XPS spectra. The data show that the Pd⁰ content of catalysts with different acid pore-widening treatments was higher than that of untreated catalysts, indicating that a larger pore size facilitates the full access of reducing substances to the interior of the microscopic pore structures, thus realizing more complete reduction. Specifically, we noticed that the Pd/(HNO₃)γ-Al₂O₃/Al had the largest content of Pd⁰, indicating that the Pd/(HNO₃)γ-Al₂O₃/Al was most fully reduced.

Figure 10 shows the conversion of nitrobenzene and the yield of aniline over granular and structured mesh-type Pd/Al₂O₃/Al catalysts as well as different acid pore-widening treatment structured mesh-type Pd/γ-Al₂O₃/Al catalysts at different liquid flow rates. The results show that the conversion of nitrobenzene increased with smaller liquid flow rate and the aniline yield increased with larger liquid flow rate under the experimental conditions (liquid flow rate 0.1 mL/min–0.5 mL/min). At the same liquid flow rate, the nitrobenzene conversion and the aniline yield of the structured mesh-type catalyst was higher than the granular catalyst, attributed to the fact that the structured mesh-type catalyst had a lower pressure drop compared to the granular catalyst, resulting in a thinner liquid film on the catalyst surface. The thinner liquid film facilitated diffusion of the H₂ from nitrobenzene solution to the catalyst surface active site [35], leading to a higher catalytic activity. As the liquid flow rate increased, the catalytic activity gap between the structured mesh-type catalyst and the granular catalyst became wider. Figure 10 also shows the conversion of nitrobenzene and the yield of aniline over different acid pore-widening treatment Pd/γ-Al₂O₃/Al catalysts at different liquid flow rates. The result shows that the activities of Pd/γ-Al₂O₃/Al catalysts with different acid pore-widening treatments were improved compared with untreated Pd/Al₂O₃/Al catalyst, due to the higher active metal loading, smaller active metal particle sizes, and increased Pd⁰ content after the acid pore-widening treatments. Notably, the Pd/(HNO₃)γ-Al₂O₃/Al exhibited the highest catalytic activity compared with Pd/(H₂C₂O₄)γ-Al₂O₃/Al and Pd/(H₂SO₄)γ-Al₂O₃/Al, owing to the largest active metal loading, the smallest active metal particle size, as well as the largest Pd⁰ content of the Pd/(HNO₃)γ-Al₂O₃/Al.

In order to further investigate the reason for different catalytic activity between granular and structured mesh-type catalysts, the pressure drop across the granular catalyst bed and structured mesh-type catalyst bed was measured and calculated.

The specific surface area of the granular and structured mesh-type catalysts can be calculated by the following Equation (4):

$$a=\frac{s}{v}$$

(4)

where $a$ denotes the specific surface area of the catalyst, $s$ denotes the surface area of the catalyst, $v$ denotes the catalyst volume.

The void ratio of the granular and structured mesh-type catalysts can be calculated by the following Equation (5):

$$\epsilon =\frac{V_b-V_c}{V_b}$$

(5)

where $\epsilon$ denotes the void ratio of the catalyst, $V_b$ denotes the bed volume of the catalyst, $V_c$ denotes the volume of the catalyst deposition.

The bed Reynolds number ${Re}^{\prime }$ can be calculated by the following Equation (6):

$${Re}^{\prime }=\frac{\rho u}{a\left(1-\epsilon \right)\mu }$$

(6)

where $\rho$ denotes the density of the nitrobenzene solution, whose value is 0.86 g/cm³, $u$ denotes the nitrobenzene solution flow rate of the fluid in the thin tub, and $\mu$ denotes the viscosity factor of the nitrobenzene solution, whose value is 2.037 × 10⁻³ Pa·s.

When ${Re}^{\prime }<2$, the pressure drop can be calculated by the Kozeny Equation (7):

$$\frac{\Delta p}{L}=K^{\prime }\frac{a^2(1-\epsilon {)}^2}{{\epsilon }^3}\mu u$$

(7)

where $\Delta p$ denotes the pressure drop, $L$ denotes the bed height, whose value is 100 mm, and $K^{\prime }$ denotes the Kozeny coefficient, whose value is 5.0 (${Re}^{\prime }<2$).

For example, with nitrobenzene solution flow rate of 0.5 mL/min (the value of $u$ is 7.37 × 10⁻³ cm/s), the bed pressure drop across the structured mesh-type catalyst and bed granular catalyst bed was calculated based on the above equations, as shown in

Table 4. Bed pressure drop and related parameters over granular and structured mesh-type catalyst.

Catalyst Layer	$\mathit{s}$ (mm2)	$\mathit{v}$ (mm3)	$\mathit{a}$ (mm−1)	${\mathit{V}}_{\mathit{b}}$ (cm3)	${\mathit{V}}_{\mathit{c}}$ (cm3)	$\mathit{\epsilon }$	${\mathit{R}\mathit{e}}^{\prime }$	$\mathsf{\Delta }\mathit{p}$
Pd/γ-Al2O3-Al-mesh-type	1.89	0.149	12.7	11.3	2.7	0.76	1.02 × 10−2	1.6
Pd/γ-Al2O3-Al-granular	2.41	0.149	13.8	11.3	7.3	0.35	3.46 × 10−3	140.9

.

Based on Table 4, it was found that the bed pressure drop of the structured mesh-type catalyst bed was 1/72 of the granular catalyst bed. Figure 11 shows that when the bed pressure drop became smaller, the pressure of liquid on the surface of structured catalysts was greater than that of granular catalysts at the same bed height. When the pressure of nitrobenzene solution on the surface of structured catalysts became greater (P₁ > P₃, P₂ > P₄), the nitrobenzene solution liquid film on the surface of structured catalysts became thinner, leading to the H₂ reaching the catalyst surface increasing and leading to an increase in the conversion rate of the reaction.

Taken together, based on the above calculations, it was demonstrated that the structured mesh-type catalyst enabled more effective diffusion of the H₂ across from nitrobenzene solution to the active sites on the surface of the catalysts, thus increasing the catalytic activity.

4. Conclusions

In this study, a novel structured mesh-type Pd/Al₂O₃/Al catalyst was developed by anodic oxidation technique. The pore structures of the γ-Al₂O₃/Al supports were designed by acid pore-widening treatments. Firstly, the pore physical properties and chemical properties of supports and catalysts were observed. Secondly, we tested nitrobenzene liquid-phase hydrogenation performance over granular and structured mesh-type Pd/Al₂O₃/Al as well as different acid pore-widening treatment structured mesh-type Pd/γ-Al₂O₃/Al catalysts. Eventually we had further investigated the deep reason behind the differences in catalytic activity between granular and structured mesh-type catalysts by calculation. The main conclusions are listed as follows.

(1)

Acid pore-widening treatments can improve the specific surface area and the pore size of the γ-Al₂O₃/Al supports. The HNO₃/γ-Al₂O₃/Al support with HNO₃ pore-widening treatment has the largest specific surface area, enlarging from 70 m²/g to 80 m²/g, and the largest pore size, enlarging from 3.7 nm to 4.6 nm.

(2)

The Pd/γ-Al₂O₃/Al catalysts with different acid pore-widening treatments contributes to the increased catalyst loading, more Pd⁰ content, and better dispersion of the metal-active particles. The Pd/(HNO₃)γ-Al₂O₃/Al prepared by support with HNO₃ pore-widening treatment has the largest active metal Pd loading, enlarging from 1.82% to 1.95%, the smallest particle size, reducing from 103 nm to 41 nm, and the largest Pd⁰ content, enlarging from 52.1% to 58.5%, resulting in the highest nitrobenzene conversion rate, increasing from 67.2% to 74.3%.

(3)

Compared to the granular catalyst, the structured mesh-type catalyst with higher porosity and more regular pore structures exerts a lower pressure drop. This facilitates diffusion of the H₂ through nitrobenzene solution to the active sites of the catalyst surface. Consequently, better conversions and yields of the structured mesh-type catalyst can be obtained in the nitrobenzene liquid-phase catalytic hydrogenation reaction.