Zirconium Phosphate Coating on Aluminum Foams by Electrophoretic Deposition for Acidic Catalysis

Abstract

The electrophoretic deposition method has been applied for the formation of an amorphous zirconium phosphate layer on the surface of open-cell aluminum foam. The aluminum foam was fully and uniformly covered by the zirconium phosphate layer with a good mechanical adherence to the support. The obtained composites were characterized by using XRD, SEM and nitrogen adsorption. The coated aluminum foams showed high catalytic activity in the dehydration of fructose to 5-hydroxymethylfurfural. This method of foam coating is much more convenient and effective than the traditional washcoating procedure, avoiding the anodization pretreatment of the foam to increase adherence.

Introduction

Solid foam reactors have attracted much attention in catalytic reactor engineering. They proffer advantageous properties such as high surface area and small diffusion length, inducing a low pressure drop owing to their open-pore cell structure. Solid foams are applied currently in environmental catalysis processes such as purification of automotive exhaust gases, reduction of NO_x exhaust gases from power stations and the catalytic combustion of volatile organic compounds.1–4 However, solid foams also have a significant potential in the application as catalyst support in multiphase processes. The high surface area allows an increase in contact of the phases in liquid–liquid or liquid–gas processes.5–6 Development of new effective methods for the coating of the catalysts on the foams should facilitate the application of these foams in the catalysis.

The most popular support material for environmental catalytic applications is the ceramic material cordierite, because of its low cost, a high mechanical strength and temperature stability, and a low thermal expansion coefficient.7 However, metallic foams have a number of advantages over ceramics, such as higher mechanical resistance, thermal conductivity, and cell densities.8 The main disadvantage of metallic monoliths compared to ceramics is low adhesion of the catalytic coating to the metallic substrate, necessitating a pretreatment to create an active surface, for example, an anodization procedure for aluminum foam with formation of an alumina layer.9 Nevertheless, the high electric conductivity of metallic foams renders them suitable for the use of electrophoretic deposition (EPD), which creates a catalytic layer on the surface of the foam. This method involves application of electrical potential to the material to attract charged particles from solution.10 The advantages of EPD are the uniformity of deposition, even for complex forms, and the level of control of the deposited layer.

EPD has been used in catalysis to create oxide layers that act as a support for subsequent catalyst coatings or as a carrier for metallic particles.11 For example, Haas-Santo et al.12 used EPD for the deposition of oxide nanoparticles (ZnO, CuO, and Al₂O₃) in microstructured metallic reactors. Vorob′eva developed catalyst carriers comprising alumina on stainless steel gauzes by use of EPD.13 Yang et al. applied EPD to prepare wire mesh monoliths for the catalytic oxidation of volatile organic compounds with an alumina catalyst.14 Nedyalkova used EPD of Co/ZnO catalysts on stainless steel plates for the steam reforming of ethanol.15

EPD has not yet been applied to the creation of an acidic catalyst layer on a structured surface. Zirconium phosphate is one of the most popular solid acidic catalysts in the world, especially in reactions involving water.16 Dehydration of carbohydrates to 5-hydroxymethyl-2-furaldehyde (HMF) has attracted increasing attention, owing to renewed interest in biomass as a source of chemicals.17 The addition of solvents that absorb HMF has been shown to significantly increase the selectivity of the process.18 The use of solid acid foams in this reaction could be an effective way of bringing the catalyst, aqueous, and organic phases into contact, thereby greatly increasing the efficiency of the process. We report a synthetic method for the preparation of an acidic zirconium phosphate layer on aluminum foams for the dehydration of fructose to HMF.

Results and discussion

Zirconium phosphate/aluminum foam composite

The EPD behavior of zirconium phosphate particles is related to their electrical properties that result, in turn, from the acidic properties of the material. For example, the preparation of zeolite membranes by EPD coating of ZSM-5 has shown that zeolite particles, owing to negative charges on the aluminumoxygen tetrahedrons, migrate towards the positively charged electrode.19 Zirconium phosphate is highly acidic owing to its large number of phosphate groups; its isoelectric point should therefore occur at a low pH, as with zeolites. On application of a constant voltage to electrodes, zirconium phosphate in slurry deposits only foam that is negatively charged. This could be a result of small amounts of phosphoric acid in the zirconium phosphate slurry. Even after several washings with water, the zirconium phosphate slurry had a pH of around 5. The presence of phosphoric acid leads to a partial protonation of oxygen in the zirconium phosphate particles, detailed in Scheme 1.

Scheme 1. Protonation of zirconium phosphate.

EPD of dried zirconium phosphate did not lead to the formation of a coating on the surface of the foam. This can be attributed to the small charged particles in the slurry being much more mobile than the large agglomerated particles formed after drying.

Photographs of Al foam before and after deposition of zirconium phosphate are shown in Figure 1. The photographs reveal the formation of a uniform white coating on the surface of the Al foam. The zirconium phosphate coating does not shrink after calcination at 673 K and ultrasound treatment showed the formation of a homogeneous layer with a good adherence to the Al foam.

Figure 1. Al foam a) before and b) after EPD deposition of zirconium phosphate.

The XRD patterns of the bulk zirconium phosphate and the aluminum foam/zirconium phosphate composite are shown in Figure 2. The XRD pattern of the calcined powder zirconium phosphate exhibits no peaks, indicating the amorphous nature of the catalyst. EPD of the catalyst on Al foam does not lead to the appearance of new peaks in addition to the signal from the Al foam support (Figure 2), indicating that EPD does not change the amorphous nature of the zirconium phosphate layer and preserves its catalytic properties.

Figure 2. XRD of bulk zirconium phosphate, zirconium phosphate coated on Al foam, and Al foam.

The relationship between EPD time and the amount of zirconium phosphate deposited at applied voltages up to 30 V is shown in Figure 3. The amount deposited increased with EPD time at voltages of 5, 15, and 30 V, with saturation after 2–3 h. The same result was observed previously with formation of zeolite Y films.20 The change in rate during deposition might have been caused by increasing deposition resistance and decreasing suspension concentration. The deposition starts on the outer surface of the foam, which is closer to the cathode. The coating deposition front moves during EPD to the anode plate because of the low resistance of the uncovered part of the foam. An increase in the applied voltage was found to achieve a large amount of deposition. For example, at an applied voltage of 30 V, the amount of zirconium phosphate deposited after 2 h was 18 % of the total mass of the foam. At an applied voltage of 5 V, the deposits totaled only 5 % of the foam mass. This implies that zirconium phosphate particles were maneuvered minimally at low voltage.

Figure 3. Dependence of deposited amount on EPD time at applied voltages of 5, 15, and 30 V.

The SEM micrograph of the synthesized bulk zirconium phosphate calcined at 673 K (Figure 4) shows globule agglomerates of small particles. The size of the agglomerates varies from 5 to 40 nm. An SEM inspection of the coatings at different magnifications shows the aluminum support to be covered totally by a zirconium phosphate layer (Figure 5 a–c). The microstructure is different to the bulk zirconium phosphate powder. In this case, the formation of a homogeneous layer is seen (Figure 5 c), which differs drastically from the globule agglomerates of the powder zirconium phosphate. Figure 5 d shows a cross section of the sample with a layer of zirconium phosphate on the aluminum surface. In the cross section, an intergrowth between the zirconium phosphate particles can be observed, thus resulting in a firm attachment to the support surface. It can also be observed that the thickness of the coating (26 μm) is high in comparison to other EPD coatings.15

Figure 4. SEM micrographs of bulk zirconium phosphate.

Figure 5. SEM micrographs of zirconium phosphate coated on Al foam at 15 V after 200 min at different magnifications.

The nitrogen adsorption–desorption isotherms and pore size distribution curves of bulk zirconium phosphate samples and zirconium phosphate/aluminum foam composites are shown in Figure 6 and Table 1. The nitrogen adsorption–desorption isotherm patterns of zirconium phosphate correspond to type IV, indicating the presence of mesopores. The horizontal hysteresis loop after p/p₀≈0.4 suggests the existence of secondary texture mesopores in the sample. In addition, the pore size distribution analysis using the Barrett–Joyner–Halenda method of the desorption isotherm of zirconium phosphate (Figure 6 b) reveals a pore size distribution with a maximum near 4 nm, which could be assigned to pores in interlayer regions between the zirconium phosphate layers or blocks.21 An increase in N₂ adsorption was observed from p/p₀≈0.8. According to the SEM data, the micrograph of the zirconium phosphate evidences agglomerates of very small particles. Adsorption of nitrogen in the range p/p₀=0.8–1.0 can, therefore, be attributed to filling of the pores between particles in agglomerates.

Figure 6. a) N₂ adsorption–desorption isotherm and b) corresponding pore size distribution curves of bulk zirconium phosphate, zirconium phosphate coated on Al foam at 15 V after 200 min, and Al foam.

Table 1. Characteristics of zirconium phosphate in bulk and coating the Al foam by EPD at 15 V.

Catalyst	EPD time [min]	ZrP/Al foam [mg]	ZrP/Al foam [wt. %]	SBET [m^2 gfoam^−1]	SBET [m^2 gZrP^−1]	Thickness [μm]
ZrP	–	–	–	–	89	–
ZrP/Al	200	320	16	21	198	26
ZrP/Al	90	204	10	17	240	14
ZrP/Al	20	72	3	9	355	9

The nitrogen adsorption–desorption isotherms of zirconium phosphate/aluminum foam composites are significantly different to that of the bulk zirconium phosphate. The parent Al foam has only a negligible adsorption capacity. The nitrogen adsorption–desorption isotherm of zirconium phosphate/aluminum foam composite does not have any hysteresis loop at p/p₀≈0.4. The pore size distribution indicates the formation of pores with sizes larger than 5 nm. This could be attributed to the ordering of the individual blocks of zirconium phosphate during EPD, which are subsequently tightly bound. The hypothesis that large macropores should be observed in the direction normal to the surface of the foam is confirmed by analysis of SEM micrographs (Figure 5).

An increase in deposition time from 20 to 200 min at 15 V leads to an increase in the amount of zirconium phosphate coating from 3 to 16 wt. % (Table 1). The surface area of the samples decreases from 21 to 9 m² g⁻¹. Recalculation of the surface area of the sample to account for the deposited zirconium phosphate indicates a gradual decrease in surface area of the zirconium phosphate over EPD time: an increased amount of zirconium phosphate coating approaches the textural properties of bulk zirconium phosphate, which has a surface area of 98 m² g⁻¹. The surface area of 10 wt. % deposited zirconium phosphate is 240 m² g⁻¹, which decreases to 198 m² g⁻¹ for 16 % wt. % deposition. EPD leads to a stabilization of the zirconium phosphate particles on the surface of the foam, thus preventing agglomeration. An increase in the amount of coated material blocks pores and the accessibility to the particles lower in the structure.

Catalytic results

The results of the catalytic performance of the coated ZrP/Al foam (EPD time=200 min, described in Table 1) and bulk zirconium phosphate in the fructose dehydration to HMF are shown in Figure 7. The activity and selectivity of 1 g of bulk and coated zirconium phosphate are compared. The conversion of fructose increased linearly over time for both samples. Notably, the activity of the foam catalyst is almost two times higher than that of the bulk catalyst, owing possibly to the surface area of the former being approximately twice as large as that of the latter (Table 1). The higher surface area and activity of the coated catalyst in comparison with the bulk catalyst indicates a higher dispersion resulting from stabilization of the catalyst particles on the foam. Similarly, a higher activity of catalyst coating, owing to highly dispersed particles on the surface, has been observed over zeolites prepared by a hydrothermal synthesis on ceramic foams in the methanol to olefin reaction,22 dimethyl ether synthesis,23 and alcohol dehydration.24

Figure 7. Fructose conversion and HMF selectivity versus time over a) bulk zirconium phosphate and b) zirconium phosphate coated on the Al foam at 15 V after 200 min.

The selectivity of the fructose transformation in HMF also increases over time. Carniti attributed this to the formation of intermediate products in the initial stages of fructose transformation to HMF.25 The 40 % selectivity of the fructose transformation over bulk zirconium phosphate corresponds to the values described earlier for crystalline zirconium phosphate phases.26 The selectivity of HMF formation is slightly lower over the coated zirconium phosphate. Oligomerization of HMF, the main side reaction, will be increased in the diffusion-limited coated layer. The zirconium phosphate layer is thicker than the agglomerate of particles in the bulk material, which might increase oligomerization in the channels of the coating.

Conclusions

A coating of zirconium phosphate on aluminum foam was produced using electrophoretic deposition. The zirconium phosphate/aluminum foam composite adheres well to the foam and coats it homogeneously. The zirconium phosphate layer on the surface of the foam has a higher surface area than the uncoated foam, resulting from stabilization of the particles on the surface of the foam, which prevents agglomeration. The coated foam has a higher catalytic activity in the dehydration of fructose than does bulk zirconium phosphate. Furthermore, the prepared coated foam provides a more convenient catalyst system with respect to catalyst separation and reuse. It can also be combined more easily with an extraction in the same reactor, leading to an intensified process with even higher conversion and selectivity.

Experimental Section

Catalyst preparation

Amorphous zirconium phosphate (ZrP) was prepared by the rapid addition of 85 % phosphoric acid (2 g, 17 mmol) in of water (20 mL) to a stirred solution of ZrOCl₂⋅8H₂O (4 g, 12 mmol) in 50 mL of water.27 The resulting suspension was stirred for 30 min at room temperature, then filtered and washed several times with large amounts of distilled water (2 L). The product was dried at 383 K and calcined at 673 K to yield bulk zirconium phosphate catalyst (2 g).

Catalyst deposition by electrophoretic coating

40 ppi (pores per inch) Duocel aluminum foams (ERG Aerospace Corp.) were used. The foam material is aluminum 6101 alloy, comprising 99.32 Al, 0.19 Mg, 0.27 Si, 0.12 Fe and 0.1 wt. % others (Cu, Mn, Zn, B). Prior to EPD the foam was cut into slabs (20×25×25 mm) with a mass of ≈2 g.

Electrophoretic coating was performed at constant voltages (5, 15, and 30 V). The coating and counter electrodes were made from 2 cm wide aluminum plates. An Al foam slab was attached to one of the plates by a wire. The electrodes were placed 1.0 cm apart in a cell with a total volume of 100 mL (Figure 8).

Figure 8. Apparatus diagram for EPD of zirconium phosphate on the Al foam.

The slurry for EPD was prepared by the addition of water (70 mL) to the filtered and washed zirconium phosphate. Mechanical stirring was maintained during EPD to prevent sedimentation of the particles. Coating time was varied from 30 to 200 min. The coated foams were dried at 383 K and calcined at 673 K for 3 h in air.

Characterization of the powder catalyst and the coated foams

The powder and the deposited catalysts were analyzed by XRD in a Rigaku Geigerflex Max/B diffractometer, using CuKα radiation from a Cu X-ray tube at 40 kV and 40 mA. SEM (Philips XL 30 ESEM-FEG) was used to explore the morphology and thickness of the coating. Cut samples were used for a cross-sectional analysis. Texture properties of the powder and deposited catalysts were determined by analysis of N₂ adsorption at 77 K in a Micromeritics ASAP 2020 surface analyzer. Before analysis, the samples were thermally treated at 573 K under vacuum to remove water.

Catalytic activity

Experiments were carried out in a 2–1 stirred autoclave working in batch mode. Zirconium phosphate powder (1 g) and water (300 mL) were placed into the autoclave, which was then purged with nitrogen. Fructose (20 g, 111 mmol) was poured into the autoclave after the temperature had been increased to 408 K and the catalytic activity subsequently monitored.

The coated foams were tested in the rotating foam stirrer reactor mode under the same conditions as for the powder catalyst (408 K, 20 g fructose, in 300 mL water). Four pieces of the Al foam coated at 15 V after 200 min contained approximately 1 g of zirconium phosphate, as found at the powder testing. The autoclave stirrer used for the powder experiment was exchanged for a holder with four pieces of the EPD-coated foam. The solid foam was used both as a catalyst support and as stirrer blades.28, 29 The agitation speed was 500 rpm for powder and foam testing. Liquid samples were taken from the autoclave periodically and analyzed using HPLC (Shimadzu) equipped with refractive index and UV-Vis detectors and a Bio-Rad Aminex HPX-87H column.

Acknowledgements

This research was performed within the framework of the CatchBio program (053.70.001).