RESTORING OF CATALYTIC MICROWAVE TREATMENTS

ALUMINA

POROUS

SPHERES

BY

Paolo Veronesi, Cristina Leonelli, Anna Bonamartini Corradi University of Modena and Reggio Emilia - Faculty of Engineering Departement of Chemistry - Via Campi 183 41100 Modena - Italy

ABSTRACT Polluted alumina porous spheres, 20-25 mm in diameter, used as catalyst in organic reactions, have been restored by fast microwave heat treatment in 2.45 GHz multi mode cavities. Using silicon carbide as an auxiliary absorber, few grams of spheres have been completely recovered after 6-8 minutes soaking at 600C, from two to four time faster when compared to conventional heating. The microwave selective and volumetric heat treatment allowed the transfer of power mainly towards the organic part, which has been rapidly decomposed, maintaining the alumina unaltered as far as pore size and interconnectivity is concerned. INTRODUCTION Porous -alumina presenting high surface area is employed as a support for catalyst, or used as it is, in many organic syntheses, in particular of chinons and hydrochinons. Reactants and reactions products often leave residua which accumulate and occlude the alumina active sites, decreasing the processes yields and eventually wasting reaction volume.. After a few months' use, most of the alumina spheres must be replaced, an operation which implies high costs due to production stops, exhaust catalyst transport and disposal and purchasing of new ones. Considering that the polluted alumina spheres are classified as special waste, disposal costs have a not negligible influence on the economic balance of the process, being about 0.25 $/kg. Catalysts regeneration has been somehow industrially implemented calcining the spheres at high temperature for 30-45 minutes in an oxygen enriched atmosphere. However, mainly due to high temperatures involved in the process, the porosity of the spheres, and thus the surface area, is negatively affected by the heat treatment. A faster treatment performed at lower temperature should avoid catalyst deterioration, provided it is

capable of completely removing the organic pollutants without leaving carbonious residua. Selective and volumetric microwave heating has been used in the course of this study to restore exhaust catalytic -alumina millimetric spheres maintaining almost unaltered their properties. EXPERIMENTAL Alumina spheres originally destined to disposal were weighed ( 2, 50, 100 and 200 g) and set on a single layer in an AL25/1700 alumina fibres cylindrical crucible of 60 mm diameter and 20 mm wall height. Materials used and geometry was maintained unaltered during the whole set of experiments, except when explicitly stated. Starting and final compositions of alumina spheres is given in table I. Table I. Alumina spheres mean weight composition Starting composition Final composition Al2O3 92 % 69 % NaO 7% 5% SiO2 1% 1% LOI negligible 25 % The alumina spheres in the alumina crucible were heat treated at 500, 600, 700C isotherms for times ranging from 120 to 1200 s. Heating was performed either by conventional resistance furnace (Nannetti PK4/O, Faenza, Italy) either by microwave multi mode furnaces (CEM MAS 7000, NC, USA and RADATHERM VPMS, Sydney, Australia). CEM furnace is temperature controlled and is equipped with a 2.45 GHz magnetron having an average output power of 950W, alternately switching on and off. On the RADATHERM furnace is installed a directional coupler which allows operating in power control, as well as continuosly varying the 2.45 GHz 2kW magnetron output power and detecting the power loss in the cavity. CEM furnace has a mode stirred cavity, while in the RADATHERM furnace the mode stirrer has been kept off during run, in order to verify its influences on homogenising the heating pattern. Microwave treatments of smaller samples ( 2g ) were conducted using silicon carbide as an auxiliary microwave absorber [1], since heat generation inside the material was not enough to compensate for power dissipations of the insulating system. A closed and more effective lining could not be adopted since oxidation reactions taking place in the spheres require a large amount of air in order not to form carbonious residua [2].

The naturally air cooled spheres were mounted on epoxy resin, polished until their larger section was evidenced and then observed either by Scanning Electron Microscopy (SEM, Philips XL40, The Netherlands) either by optical CCD camera (Mitsubishi Microwatcher VS-30H, Tokyo, Japan). The acquired digital images were analysed using UTHSCSA Image Tool software (UTHSCSA, S. Antonio, Texas) in order to measure the thickness of the reacted layer and to estimate the pore size. RESULTS The starting spheres, contaminated by organic reactants, present a reddish coloration in the whole body, while after heat treatment it is possible to distinguish between an outer reacted white region and an inner not completely reacted black one, as shown in figure 1.

Figure 1 : Polluted (left) and partially recovered (right) alumina spheres Conventional heat treatment: The spheres were treated at 500, 600, 700C for different periods, leading to full recovery of the catalysts in times varying from 480 to 1200 s for 2g samples. Figure 2 shows the evolution of reacted layer of the sectioned spheres with time and temperature, while in figure 3 are reported the curves of the reacted depth as a function of time.

500C

a)

600C

b)

700C

c)

Figure 2 - Cross section of the conventionally heat-treated samples, (a) 500C for 300 s, 600 s and 900 s, (b) 600C for 240 s, 360 s, 480 s, 600 s, 900 s and 1200 s, (c) 700C for 120 s, 240 s, 360 s, 480 s and 600 s.

1.20

1.00

0.80

d (mm)

0.60

0.40

0.20

C 500C C 600C C 700C

0.00 0 200 400 Time (s) 600 800 1000

Figure 3 - Recovery curves for conventionally heated samples Microwave mode stirred heat treatment: The spheres were treated at 500, 600, 700C for different periods, leading to full recovery of the catalysts in times varying from a half to a quarter of the conventional treatments for a 2 g sample. Figure 4 shows the experimental results, while in figure 5 are reported the curves of the reacted depth as a function of time.

500C

a)

600C

b)

700C

c)

Figure 4 - Cross section of the microwave heat-treated samples(a) 500C for 300 s, 600 s and 900 s, (b) 600C for 120 s, 240 s, 360 s and 480 s (c) 700C for 120 s, 240 s and 360 s.

1.40 1.20 1.00 d (mm) 0.80 0.60 0.40 0.20 0.00 0 200 400 Time (s) 600 800 1000 MW 500C MW 600C MW 700C

Figure 5 - Recovery curves for microwave heated samples Non-stirred microwave heat treatment: The spheres were treated at 600C, temperature chosen as a compromise between fast recovery and alumina spheres properties maintaining. Figure 6 shows the experimental results, evidencing a less pronounced homogeneity compared to the samples treated in the mode stirred cavity. Treatment times are further reduced due to sample positioning in a high field region, as qualitatively confirmed by heating patterns obtained by thermal paper impression [2].
1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0 100 200 Time (s) 300 400 d (mm)

600C

Figure 6 - Cross section and recovery curve for the non stirred microwave samples heated at 600C for 120 s, 240 s and 360 s.

DISCUSSION From the elaboration of reacted layer thickness data it was possible to evidence a dependence of the reacted distance from the square of the soaking time, with correlation factor R2 ranging from 9.988 to 9.999. Temperature increases, enhancing diffusion, lead to complete recovery of the catalyst in shorter times. This effect is more pronounced passing from 600 to 700C rather than from 500 to 600C. These are the only similarities between conventional and microwave treatment. In fact, for temperatures lower than 600C, -alumina can be considered almost transparent to microwaves, if compared to the organic part polluting the spheres [3]. This implies a preferential energy transfer towards the organic part, while alumina is almost unaffected by the irradiation. This leads to a heat transfer extended to the whole spheres' body and much faster, if compared to the slower transmission mechanisms involved in the conventional heating. Figures 7a, 7b, 7c show a comparison between conventional and microwave treatments performed at the same temperature, respectively of 500, 600, 700C.
1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0 200 400 600 800 time (s) 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0 500 time (s) 1000

d (mm)

Conv MW

d (mm)

Conv MW

1000

1500

a)
1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0 200

b)
1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0 100 200

d (mm)

d (mm)

Stirrer OFF Stirrer ON 300 Time (s) 400 500 600

Conv MW

400 time (s)

600

c)

d)

Figure 7 - Samples treated conventionally and by microwaves at a)500C, b)600C, c)700C, d)600C with and without stirring Tests performed on larger samples, weighing 50, 100 and 200 g, not using auxiliary absorbers, evidenced an initial microwave power absorption of about 3540% of the forward power, constantly decreasing down to 15-20% as the organic part was removed. It was not possible to achieve total removal since the experimental geometry, i.e. an alumina crucible having walls 13 cm high, did not allow complete oxidation of the inner layers, due to depletion of oxygen during

the treatment. The same procedure, in conventional furnace, lead to the same problems; however the recovered part, on the surface, weighed 45% less than the part recovered by microwave treatments. Loss power measurements performed in presence of the auxiliary absorbers are not considered meaningful since more than 95% of the forward power is continuosly absorbed by the silicon carbide elements, masking the sample's behaviour. As far as sample degradation is concerned, scanning electron microscopy evidenced that relatively high temperature (900C) heat treatments lasting even only 10 minutes lead to densification of the spheres, reducing the porosity and thus the surface area. Figure 8 shows a comparison between a sphere untreated, heated for 600 s at 600C, heated for 600 s at 900C.

Figure 8 - SEM micrographs of (a) untreated sample, (b) 600 s heat treated sample at 600C, (c) 600 s heat treated sample at 900C CONCLUSIONS Millimetric -alumina porous spheres used as catalyst can be successfully recovered by fast microwave heat treatment at 2.45 GHz. The short times involved in the process, as well as the relatively low temperatures required to achieve complete recovery, avoid the premature deterioration of the spheres. From the experimental results it emerged that an isothermal treatment at 600C for 6-8 minutes is enough to guarantee full restoring of few grams of the porous spheres. The heat treatment of large samples, either by microwave or conventionally, requires devices to ensure contact between air and all the spheres, otherwise, due to the low temperatures and short times, it is impossible to avoid carbonious residua formation. This could be simply acquainted by keeping the spheres moving and revolving while heating, or continuously feeding the microwave applicator with small quantities of material. At the moment two small portable prototype systems are under investigation: a continuos furnace fed by traslating plates and an hybrid discontinuous system conveying hot air directly in the centre of the material under treatment. The choice and subsequent implementation of one of the two systems is strictly connected to economic considerations, especially as

far as the frequency of operation and the amount of material to be recovered are concerned. REFERENCES [1] W.H. Sutton, "Microwave processing of ceramic materials", Cer.Bull., 68 [2] 376-86 (1989) [2] K.Y.Lee and E.D.Case, "Microwave binder burn-out for batch processing of Al2O3 Al2O3/SiC platelet, and Al2O3/ZrO2 particle powder compacts", Cer. Trans. 80 539-46 (1997) [3] Bradshaw S., Delport S. and E.Van Wyk, "Qualitative measurement of heating uniformity in a multimode microwave cavity", Journal of Microwave power and electromagnetic energy, 32 [2] 87-95 (1997) [4] E.H. Moore, D.E. Clark and R. Hutcheon, "Polymethil methacrilate binder emoval from an alumina compact: microwave versus conventional heating", pp. 341-46 in Microwave processing of materials III, R.L. Beatty, W.H. Sutton and M.F. Iskander, MRS, Pittsburgh, 1992