Process for preparing phthalic anhydride

Abstract

The present invention relates to a process for preparing phthalic anhydride by gas phase oxidation of aromatic hydrocarbons, in which a gas stream comprising at least one aromatic hydrocarbon and molecular oxygen is passed continuously over a thermostatted catalyst and the supply of the at least one aromatic hydrocarbon to the catalyst is temporarily interrupted after putting the catalyst on stream.

Claims

1. A process for preparing phthalic anhydride by gas phase oxidation of o-xylene and/or napthalene, the process comprising: continuously passing a gas stream comprising at least one aromatic hydrocarbon selected from the group consisting of o-xylenes and naphthalene, and molecular oxygen over a thermostatted catalyst in at least one reactor tube, wherein said at least one reactor tube is surrounded by a heat carrier medium, interrupting the supply of the at least one aromatic hydrocarbon to the catalyst for a period after putting the catalyst on stream and resuming the supply of the at least one aromatic hydrocarbon to the catalyst at the end of the period; wherein the temperature of the heat carrier medium is set to a higher value at the end of the period than before the interruption of the supply of the aromatic hydrocarbon to the catalyst.

2. The process according to claim 1, wherein the length of the period is in the range of 30 minutes to 200 days.

3. The process according to claim 1, wherein a hydrocarbon-free gas stream is passed over the catalyst during the period.

4. The process according to claim 3, wherein the hydrocarbon-free gas stream passed over the catalyst comprises between 0 and 50% by volume of oxygen and otherwise comprises air, nitrogen and/or noble gases.

5. The process according to claim 1, wherein no gas stream is passed over the catalyst during the period.

6. The process according to claim 1, wherein the gas flow rate is set to a lower value at the end of the period than before the interruption of the supply of the aromatic hydrocarbon to the catalyst.

7. The process according to claim 1, wherein the loading of the gas stream with aromatic hydrocarbon is set to a lower value at the end of the period than before the interruption of the supply of the aromatic hydrocarbon to the catalyst.

8. The process according to claim 1, wherein the temperature of the heat carrier medium is set to a 1 to 5° C. higher value at the end of the period than before the interruption of the supply of the aromatic hydrocarbon to the catalyst.

9. The process according to claim 1, wherein the gas flow rate is set to a 5 to 15% lower value at the end of the period than before the interruption of the supply of the aromatic hydrocarbon to the catalyst.

10. The process according to claim 1, wherein the loading of the gas stream with aromatic hydrocarbon is set to a 5 to 30% lower value at the end of the period than before the interruption of the supply of the aromatic hydrocarbon to the catalyst.

11. The process according to claim 1, wherein the interrupting occurs by replacing the reactant gas with nitrogen for two hours to 52 days.

12. The process according to claim 1, wherein the interrupting occurs by purging with air for 2 minutes, then stopping the flow of gas for 12 hours to three days.

13. A process for preparing phthalic anhydride by gas phase oxidation of o-xylene and/or napthalene, the process comprising: continuously passing a gas stream comprising at least one aromatic hydrocarbon selected from the group consisting of o-xylenes and naphthalene, and molecular oxygen over a thermostatted catalyst in at least one reactor tube, wherein said at least one reactor tube is surrounded by a heat carrier medium, interrupting the supply of the at least one aromatic hydrocarbon to the catalyst for a period after putting the catalyst on stream and resuming the supply of the at least one aromatic hydrocarbon to the catalyst at the end of the period; wherein the temperature of the heat carrier medium is set to a 1 to 5° C. higher value at the end of the period than before the interruption of the supply of the aromatic hydrocarbon to the catalyst; wherein the gas flow rate is set to a 5 to 15% lower value at the end of the period than before the interruption of the supply of the aromatic hydrocarbon to the catalyst; and wherein the interrupting occurs by purging with air for 2 minutes, then stopping the flow of gas for 12 hours to three days.

Description

EXAMPLES

Example 1 (Inventive)

(1) Catalyst Zone 1 (CZ1) (Vanadium Antimonate as V and Sb Source):

(2) Preparation of the Vanadium Antimonate:

(3) 2284.1 g of vanadium pentoxide and 1462 g of antimony trioxide (Antraco ACC-BS, approx. 4% valentinite and 96% senarmontite; Sb.sub.2O.sub.3≧99.8% by weight: As≦800 ppm by weight, Pb≦800 ppm by weight, Fe≦30 ppm by weight, mean particle size=1.4 μm) were suspended in 5.6 of demineralized water and the suspension was stirred under reflux for 15 hours. Thereafter, the suspension was cooled to 80° C. and dried by means of spray drying. The inlet temperature was 340° C., the exit temperature 110° C. The spray powder thus obtained had a BET surface area of 89 m.sup.2/g and had a vanadium content of 32% by weight and an antimony content of 30% by weight. The product had the following crystalline constituents: valentinite (ICPDS: 11-0689): approx. 3%; senarmontite (ICPDS: 43-1071): approx. 2%; vanadium antimonate (ICPDS: 81-1219): approx. 95%. The mean crystal size of the vanadium antimonate was approx. 9 nm, Suspension mixing and coating:

(4) 2 kg of steatite rings (magnesium silicate) having dimensions of 7 mm×7 mm×4 mm were coated in a fluidized bed apparatus with 752 g of a suspension of 4.4 g of cesium carbonate, 413.3 g of titanium dioxide (Fuji TA 100 CT; anatase, BET surface area 27 m.sup.2/g), 222.5 g of titanium dioxide (Fuji TA 100; anatase, BET surface area 7 m.sup.2/g), 86.9 g of vanadium antimonate (as prepared above), 1870.1 g of demineralized water and 76.7 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(5) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 8.4%. The analyzed composition of the active material consisted of 7.1% V.sub.2O.sub.5, 4.5% Sb.sub.2O.sub.3, 0.50% Cs, remainder TiO.sub.2.

(6) Catalyst Zone 2 (CZ2) (Vanadium Pentoxide and Antimony Trioxide, Respectively, as V and Sb Sources):

(7) 2 kg of steatite rings (magnesium silicate) having dimensions of 7 mm×7 mm n4 mm were coated in a fluidized bed apparatus with 920 g of a suspension of 3.0 g of cesium carbonate, 446.9 g of titanium dioxide (Fuji TA 100 C; anatase, BET surface area 20 m.sup.2/g), 133.5 g of titanium dioxide (Fuji TA 100; anatase, BET surface area 7 m.sup.2/g), 45.4 g of vanadium pentoxide, 11.6 g of antimony trioxide, 1660.1 g of demineralized water and 104.5 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(8) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 10.0%. The analyzed composition of the active material consisted of 7.1% V.sub.2O.sub.5, 1.8% Sb.sub.2O.sub.3, 0.38% Co. remainder TiO.sub.2.

(9) Catalyst Zone 3 (CZ3) (Vanadium Pentoxide and Antimony Trioxide, Respectively, as V and Sb Sources):

(10) 2 kg of steatite rings (magnesium silicate) having dimensions of 7 mm×7 mm×4 mm were coated in a fluidized bed apparatus with 750 g of a suspension of 2.33 g of cesium carbonate, 468.7 g of titanium dioxide (Fuji TA 100 C; anatase, BET surface area 20 m.sup.2/g), 76.3 g of titanium dioxide (Fuji TA 100; anatase, BET surface area 7 m.sup.2/g), 48.7 g of vanadium pentoxide, 16.7 g of antimony trioxide, 1588.0 g of demineralized water and 85.2 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(11) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 8.4%. The analyzed composition of the active material consisted of 7.95% V.sub.2O.sub.5, 2.7% Sb.sub.2O.sub.3, 0.31% Cs, remainder TiO.sub.2.

(12) Catalyst Zone 4 (CZ4) (Vanadium Pentoxide and Antimony Trioxide, Respectively, as V and Sb Sources):

(13) 2 kg of steatite rings (magnesium silicate) having dimensions of 7 mm×7 mm×4 mm were coated in a fluidized bed apparatus with 760 g of a suspension of 1.7 g of cesium carbonate, 370.1 g of titanium dioxide (Fuji TA 100 CT; anatase, BET surface area 27 m.sup.2/g), 158.6 g of titanium dioxide (Fuji TA 100; anatase, BET surface area 7 m.sup.2/g), 67.3 g of vanadium pentoxide, 14.8 g of antimony trioxide, 1587.9 g of demineralized water and 86.3 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(14) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 8.7%. The analyzed composition of the active material consisted of 11% V.sub.2O.sub.5, 2.4% Sb.sub.2O.sub.3, 0.22% Co. remainder TiO.sub.2.

(15) Catalyst Zone 5 (CZ5) (Vanadium Pentoxide and Antimony Trioxide, Respectively, as V and Sb Sources):

(16) 2 kg of steatite rings (magnesium silicate) having dimensions of 7 mm×7 mm×4 mm were coated in a fluidized bed apparatus with 850 g of a suspension of 389.8 g of titanium dioxide (Fuji TA 100 CT; anatase, BET surface area 27 m.sup.2/g), 97.5 g of titanium dioxide (Fuji TA 100; anatase, BET surface area 7 m.sup.2/g), 122.4 g of vanadium pentoxide, 1587.9 g of demineralized water and 96.5 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(17) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 9.1%. The analyzed composition of the active material consisted of 20% V.sub.2O.sub.5, 0.38% P, remainder TiO.sub.2.

(18) The catalytic oxidation of o-xylene to phthalic anhydride was performed in a salt bath-cooled tubular reactor having an internal tubular diameter of 25 mm. From reactor inlet to reactor outlet, 80 cm of CZ1, 60 cm of CZ2, 70 cm of CZ3, 50 cm of CZ4 and 60 cm of CZ5 were introduced into an iron tube of length 3.5 m with internal width 25 mm. For temperature regulation, the iron tube was surrounded by a salt melt; a 4 mm external diameter thermowell with installed tensile element served for catalyst temperature measurement.

(19) 4.0 m.sup.3 (STP) of air per hour flowed through the tube from the top downward with loadings of 99 to 99.4% by weight o-xylene of 30 to 100 g/m.sup.3 (STP). After a run time of 172 days, at a salt bath temperature of 346° C., an air flow rate of 4 m.sup.3 (STP)/h and an o-xylene loading of 100 g/m.sup.3 (STP), the supply of the reactant gas to the catalyst was interrupted for 6 hours and replaced by nitrogen (reactor inlet pressure=260 mbar). After the 6 hours, the catalyst was charged again under conditions identical to before the shutdown with o-xylene-laden air, i.e. a salt bath temperature of 346° C., an air flow rate of 4 m.sup.3 (STP)/h and an o-xylene loading of 100 g/m.sup.3 (STP). After four further days, the product gas composition was analyzed (see Table 1).

(20) TABLE-US-00001 TABLE 1 Before the After the shutdown shutdown Days since shutdown 0 4 Air rate [m.sup.3 (STP)/h] 4.0 4.0 Loading [g.sub.o-X/m.sup.3 (STP)] 100 100 Salt bath temperature [° C.] 346 346 o-Xylene [% by wt.] 0.071 0.070 Phthalide [% by wt.] 0.119 0.100

(21) It has been found that, under identical conditions, the product quality after the interruption of the supply of the aromatic hydrocarbon to the catalyst was improved, recognizable by the reduction in the content of phthalide, an underoxidation product, at constant o-xylene concentration in the product.

Example 2 (Inventive)

(22) 4.0 m.sup.3 (STP) of air per hour flowed through a tube which corresponds to that described in example 1 and which has also been filled with the same catalyst bed as described in example 1 from the top downward with loadings of 99 to 99.4% by weight o-xylene of 30 to 100 g/m.sup.3 (STP). After a run time of 249 days, at a salt bath temperature of 351° C., an air flow rate of 4 m.sup.3 (STP)/h and an o-xylene loading of 100 g/m.sup.3 (STP), the supply of the reactant gas to the catalyst was interrupted for 52 days and replaced by nitrogen (reactor inlet pressure=260 mbar). The catalyst was subsequently started up again at an elevated salt bath temperature and a reduced o-xylene loading. Within two weeks, the settings as before the shutdown were attained and the product gas composition was analyzed (see Table 2).

(23) TABLE-US-00002 TABLE 2 Before the After the shutdown Restart shutdown Days since shutdown 0 52 66 Air rate [m.sup.3 (STP)/h] 4.0 4.0 4.0 Loading [g.sub.o-X/m.sup.3 (STP)] 100 70 100 Salt bath temperature [° C.] 348 352 347.7 o-Xylene [% by wt.] 0.021 0.012 Phthalide [% by wt.] 0.054 0.032

Example 3 (Inventive)

(24) The catalytic oxidation of o-xylene to phthalic anhydride was performed in a salt bath-cooled tubular reactor having an internal tubular diameter of 25 mm. From reactor inlet to reactor outlet, 90 cm of CZ1, 60 cm of CZ2, 70 cm of CZ3, 50 cm of CZ4 and 60 cm of CZ5 were introduced into an iron tube of length 3.5 m with internal width 25 mm. For temperature regulation, the iron tube was surrounded by a salt melt; a 4 mm external diameter thermowell with installed tensile element served for catalyst temperature measurement.

(25) 4.0 m.sup.3 (STP) of air per hour flowed through the tube from the top downward with loadings of 99 to 99.4% by weight o-xylene of 30 to 100 g/m.sup.3 (STP). After a run time of 50 days, at a salt bath temperature of 360° C., an air flow rate of 4 m.sup.3 (STP)/h and an o-xylene loading of 83 g/m.sup.3 (STP), the supply of the reactant gas to the catalyst was interrupted. The reactor was purged with air for about 2 minutes and then the gas supply was stopped entirely for 3 days. The catalyst was subsequently started up again at an elevated salt bath temperature, a reduced o-xylene loading and a reduced air flow rate. Within 7 days, the settings as before the shutdown were attained and the by-products in the product gas composition were analyzed (see Table 3).

(26) TABLE-US-00003 TABLE 3 Before the After the shutdown Restart shutdown Days since shutdown 0 3 10 Air rate [m.sup.3 (STP)/h] 4.0 3.8 4.0 Loading [g.sub.o-X/m.sup.3 (STP)] 83 67 83 Salt bath temperature [° C.] 360 363 360 o-Xylene [% by wt.] 0.011 0.009 Phthalide [% by wt.] 0.025 0.019

Example 4 (Inventive)

(27) 4.0 m.sup.3 (STP) of air per hour with loadings of 99 to 99.4% by weight o-xylene of 30 to 100 g/m.sup.3 (STP) flowed from the top downward through a tube which corresponds to that described in example 3 and which has also been filled with the same catalyst bed as described in example 3. After a run time of 12 days, at a salt bath temperature of 367° C., an air flow rate of 4 m.sup.3 (STP)/h and an o-xylene loading of 58 g/m.sup.3 (STP), the supply of the reactant gas to the catalyst was interrupted. The reactor was purged with air for about 2 minutes and then the gas supply was stopped entirely for 12 hours. The catalyst was subsequently started up again at an elevated salt bath temperature, an equal o-xylene loading and a reduced air flow rate. Within 8 days, the settings as before the shutdown were attained and the by-products in the product gas composition were analyzed (see Table 4).

(28) TABLE-US-00004 TABLE 4 Before the After the shutdown Restart shutdown Days since shutdown 0 0 8 Air rate [m.sup.3 (STP)/h] 4.0 3.8 4.0 Loading [g.sub.o-X/m.sup.3 (STP)] 58 58 58 Salt bath temperature [° C.] 367 370 367 o-Xylene [% by wt.] 0.093 0.046 Phthalide [% by wt.] 0.199 0.107

Example 5 (Inventive)

(29) 4.0 m.sup.3 (STP) of air per hour with loadings of 99 to 99.4% by weight o-xylene of 30 to 100 g/m.sup.3 (STP) flowed from the top downward through a tube which corresponds to that described in example 3 and which has also been filled with the same catalyst bed as described in example 3. After a run time of 89 days, at a salt bath temperature of 367° C., an air flow rate of 4 m.sup.3 (STP)/h and an o-xylene loading of 58 g/m.sup.3 (STP), the supply of the reactant gas to the catalyst was interrupted. The reactor was purged with air for about 2 minutes and then the gas supply was stopped entirely for 12 hours. The catalyst was subsequently started up again at the same salt bath temperature, a reduced o-xylene loading and a reduced air flow rate. Within 5 days, the settings as before the shutdown were attained and the by-products in the product gas composition were analyzed (see Table 5).

(30) TABLE-US-00005 TABLE 5 Before the After the shutdown Restart shutdown Days since shutdown 0 0 5 Air rate [m.sup.3 (STP)/h] 4.0 3.5 4.0 Loading [g.sub.o-X/m.sup.3 (STP)] 92 87 92 Salt bath temperature [° C.] 348 348 348 o-Xylene [% by wt.] 0.047 0.037 Phthalide [% by wt.] 0.071 0.062

Example 6 (Noninventive)

(31) The catalytic oxidation of o-xylene to phthalic anhydride was performed in a salt bath-cooled industrial tubular reactor having an internal tubular diameter of 25 mm. From reactor inlet to reactor outlet, 81 cm of CZ1, 70 cm of CZ2, 85 cm of CZ3, 50 cm of CZ4 and 60 cm of CZ5 were introduced into an iron tube of length 4.0 m with internal width 25 mm. The iron tube was surrounded by a salt melt for temperature regulation.

(32) 3.8 m.sup.3 (STP) of air per hour with loadings of 98 to 99% by weight o-xylene of 30 to 100 g/m.sup.3 (STP) flowed through the tube from the top downward. After a run time of 213 days, the salt bath temperature of 345° C., an air flow rate of 3.8 m.sup.3 (STP)/h and an o-xylene loading of 95.4 g/m.sup.3 (STP) were kept constant for 9 days and then the by-products in the product gas composition were analyzed (see Table 6).

(33) TABLE-US-00006 TABLE 6 Run time [days] 213 222 Air rate [m.sup.3 (STP)/h] 3.8 3.8 Loading [g.sub.o-X/m.sup.3 (STP)] 95.4 95.4 Salt bath temperature [° C.] 345 345 o-Xylene [% by wt.] 0.08 0.08 Phthalide [% by wt.] 0.08 0.08

Example 7 (Noninventive)

(34) 3.8 m.sup.3 (STP) of air per hour with loadings of 98 to 99% by weight o-xylene of 30 to 100 g/m.sup.3 (STP) flowed from the top downward through a tube which corresponds to that described in example 6 and which has also been filled with the same catalyst bed as described in example 6. After a run time of 273 days, the salt bath temperature of 345° C., an air flow rate of 3.8 m.sup.3 (STP)/h and an o-xylene loading of 95.9 g/m.sup.3 (STP) were kept constant for 21 days, and the by-products in the product gas composition were analyzed (see Table 7).

(35) TABLE-US-00007 TABLE 7 Run time [days] 273 294 Air rate [m.sup.3 (STP)/h] 3.8 3.8 Loading [g.sub.o-X/m.sup.3 (STP)] 95.9 95.9 Salt bath temperature [° C.] 345 345 o-Xylene [% by wt.] 0.06 0.06 Phthalide [% by wt.] 0.06 0.07

Example 8 (Inventive)

(36) Catalyst Zone 1 (CZ1)

(37) Suspension Mixing and Coating:

(38) 2 kg of steatite rings (magnesium silicate) in the form of rings having dimensions of 8 mm×6 mm×5 mm were coated in a fluidized bed apparatus with 855 g of a suspension of 6.9 g of cesium carbonate, 574.6 g of titanium dioxide (Fuji TA 100 C; anatase, BET surface area 20 m.sup.2/g), 30.6 g of vanadium pentoxide, 1.75 g of niobium pentoxide, 1588.0 g of demineralized water and 97.1 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(39) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 8.9%. The analyzed composition of the active material consisted of 5% V.sub.2O.sub.5, 0.2% Nb.sub.2O.sub.5, 0.92% Cs, remainder TiO.sub.2.

(40) Catalyst Zone 2 (CZ2):

(41) 2 kg of steatite rings (magnesium silicate) in the form of rings having dimensions of 8 mm×6 mm×5 mm were coated in a fluidized bed apparatus with 960 g of a suspension of 5.0 g of cesium sulfate, 394.4 g of titanium dioxide (Fuji TA 100 CT; anatase, BET surface area 27 m.sup.2/g), 169.5 g of titanium dioxide (Fuji TA 100; anatase, BET surface area 7 m.sup.2/g), 43.3 g of vanadium pentoxide, 1.75 g of niobium pentoxide, 0.6 g of potassium sulfate, 0.7 g of ammonium dihydrogenphosphate, 1584.8 g of demineralized water and 109 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(42) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 9.0%. The analyzed composition of the active material consisted of 7% V.sub.2O.sub.5, 0.2% Nb.sub.2O.sub.5, 0.6% Cs, 0.04% K, 0.03% P, remainder TiO.sub.2.

(43) Catalyst Zone 3 (CZ3):

(44) 2 kg of steatite rings (magnesium silicate) in the form of rings having dimensions of 8 mm×6 mm×5 mm were coated in a fluidized bed apparatus with 950 g of a suspension of 3.5 g of cesium sulfate, 395.6 g of titanium dioxide (Fuji TA 100 CT; anatase, BET surface area 27 m.sup.2/g), 169.5 g of titanium dioxide (Fuji TA 100; anatase, BET surface area 7 m.sup.2/g), 42.9 g of vanadium pentoxide, 1.75 g of niobium pentoxide, 0.6 g of potassium sulfate, 0.7 g of ammonium dihydrogenphosphate, 1585.6 g of demineralized water and 107.9 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(45) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 9.0%. The analyzed composition of the active material consisted of 7% V.sub.2O.sub.5, 0.2% Nb.sub.2O.sub.5, 0.35% Cs, 0.04% K, 0.03% P, remainder TiO.sub.2.

(46) Catalyst Zone 4 (CZ4):

(47) 2 kg of steatite rings (magnesium silicate) in the form of rings having dimensions of 8 mm×6 mm×5 mm were coated in a fluidized bed apparatus with 780 g of a suspension of 442.78 g of titanium dioxide (Fuji TA 100 C; anatase, BET surface area 20 m.sup.2/g), 111.3 g of vanadium pentoxide, 1.7 g of tungsten trioxide, 3.8 g of ammonium dihydrogenphosphate, 1443.6 g of demineralized water and 88.6 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(48) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 8.0%. The analyzed composition of the active material consisted of 20% V.sub.2O.sub.5, 0.18% P, 0.24% W, remainder TiO.sub.2.

(49) The catalytic oxidation of o-xylene/naphthalene to phthalic anhydride was performed in a salt bath-cooled tubular reactor having an internal tubular diameter of 25 mm. From reactor inlet to reactor outlet, 80 cm of CZ1, 80 cm of CZ2, 90 cm of CZ3 and 70 cm of CZ4 were introduced into an iron tube of length 3.5 m with internal width 25 mm. For temperature regulation, the iron tube was surrounded by a salt melt; a 4 mm external diameter thermowell with installed tensile element served for catalyst temperature measurement.

(50) 4.0 m.sup.3 (STP) of air per hour with loadings of 99 to 99.4% by weight o-xylene of 0 to 40 g/m.sup.3 (STP) and technical naphthalene of 37 to 42 g/m.sup.3 (STP) flowed through the tube from the top downward. After a run time of 5 days, at a salt bath temperature of 377° C., an air flow rate of 4 (STP)/h, an o-xylene loading of 10 g/m.sup.3 (STP) and a naphthalene loading of 41.2 g/m.sup.3 (STP), the supply of the reactant gas to the catalyst was interrupted for 2 hours and replaced by nitrogen. After the 2 hours, the catalyst was charged again with o-xylene- and naphthalene-laden air under conditions identical to those before the shutdown, i.e. a salt bath temperature of 377° C., an air flow rate of 4 m.sup.3 (STP)/h, an o-xylene loading of 10 g/m.sup.3 (STP) and a naphthalene loading of 41.2 g/m.sup.3 (STP). After a further day, the product gas composition was analyzed (see Table 8).

(51) TABLE-US-00008 TABLE 8 Before the After the shutdown shutdown Days since shutdown 0 1 Air rate [m.sup.3 (STP)/h] 4.0 4.0 Loading [g.sub.o-X/m.sup.3 (STP)] 10 10 Loading [g.sub.naphthalene/m.sup.3 (STP)] 41.2 41.2 Salt bath temperature [° C.] 377 377 o-Xylene [% by wt.] 0.035 0.008 Phthalide [% by wt.] 0.044 0.013 Naphthoquinone [% by wt.] 2.145 1.157

(52) It has been found that, under identical conditions, the product quality after the interruption of the supply of the aromatic hydrocarbon to the catalyst has been improved, noticeable by the reduction in the content of phthalide and naphthoquinone, which are underoxidation products, with equal o-xylene and naphthalene concentration in the product.

Example 9 (Inventive)

(53) Catalyst Zone 1 (CZ1)

(54) Suspension Mixing and Coating:

(55) 2 kg of steatite rings (magnesium silicate) in the form of rings having dimensions of 8 mm×6 mm×5 mm were coated in a fluidized bed apparatus with 870 g of a suspension of 6.92 g of cesium carbonate, 562.34 g of titanium dioxide (Fuji TA 100 C; anatase, BET surface area 20 m.sup.2/g), 42.86 g of vanadium pentoxide, 1.75 g of niobium pentoxide, 1587.96 g of demineralized water and 98.8 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(56) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 8.9%. The analyzed composition of the active material consisted of 4.62% V.sub.2O.sub.5, 0.28% Nb.sub.2O.sub.5, 0.99% Cs, remainder TiO.sub.2.

(57) Catalyst Zone 2 (CZ2):

(58) 2 kg of steatite rings (magnesium silicate) in the form of rings having dimensions of 8 mm×6 mm×5 mm were coated in a fluidized bed apparatus with 970 g of a suspension of 5.3 g of cesium sulfate, 562.3 g of titanium dioxide (Fuji TA 100 C; anatase, BET surface area 20 m.sup.2/g), 39.0 g of vanadium pentoxide, 1.75 g of niobium pentoxide, 1441.6 g of demineralized water and 110.2 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(59) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 9.0%. The analyzed composition of the active material consisted of 7% V.sub.2O.sub.5, 0.2% Nb.sub.2O.sub.5, 0.67% Cs, remainder TiO.sub.2.

(60) Catalyst Zone 3 (CZ3):

(61) 2 kg of steatite rings (magnesium silicate) in the form of rings having dimensions of 8 mm×6 mm×5 mm were coated in a fluidized bed apparatus with 900 g of a suspension of 3.5 g of cesium sulfate, 565.5 g of titanium dioxide (Fuji TA 100 C; anatase, BET surface area 20 m.sup.2/g), 42.9 g of vanadium pentoxide, 1.75 g of niobium pentoxide, 1441.6 g of demineralized water and 102.2 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(62) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 9.0%. The analyzed composition of the active material consisted of 7% V.sub.2O.sub.5, 0.2% Nb.sub.2O.sub.5, 0.4% Cs, remainder TiO.sub.2.

(63) Catalyst Zone 4 (CZ4):

(64) 2 kg of steatite rings (magnesium silicate) in the form of rings having dimensions of 8 mm×6 mm×5 mm were coated in a fluidized bed apparatus with 865 g of a suspension of 198.3 g of titanium dioxide (Fuji TA 100 CT; anatase, BET surface area 20 m.sup.2/g), 368.3 g of titanium dioxide (Fuji TA 100; anatase, BET surface area 7 m.sup.2/g), 42.9 g of vanadium pentoxide, 1.75 g of niobium pentoxide, 5.1 g of ammonium dihydrogenphosphate, 1587.9 g of demineralized water and 98.3 g of organic binder (copolymer of vinyl acetate and vinyl laurate in the form of a 50% by weight aqueous dispersion).

(65) After calcination of the catalyst at 450° C. for one hour, the active material applied to the steatite rings was 9.6%. The analyzed composition of the active material consisted of 7% V.sub.2O.sub.5, 0.2% Nb.sub.2O.sub.5, 0.22% P (introduced into the suspension as dihydrogenphosphate), remainder TiO.sub.2.

(66) The catalytic oxidation of o-xylene/naphthalene to phthalic anhydride was performed in a salt bath-cooled tubular reactor having an internal tubular diameter of 25 mm. From reactor inlet to reactor outlet, 80 cm of CZ1, 80 cm of CZ2, 90 cm of CZ3 and 90 cm of CZ4 were introduced into an iron tube of length 3.5 m with internal width 25 mm. For temperature regulation, the iron tube was surrounded by a salt melt; a 4 mm external diameter thermowell with installed tensile element served for catalyst temperature measurement.

(67) The tube was passed from the top downward with 4.0 m.sup.3 (STP) of air per hour with loadings of 99 to 99.4% by weight o-xylene of 0 to 25 g/m.sup.3 (STP) and technical naphthalene of 38 to 41 g/m.sup.3 (STP). After a run time of 4 days, at a salt bath temperature of 380° C., an air flow rate of 4 m.sup.3 (STP)/h, an o-xylene loading of 0 g/m.sup.3 (STP) and a naphthalene loading of 40 g/m.sup.3 (STP), the supply of the reactant gas to the catalyst was interrupted for 4 hours and replaced by nitrogen. After the 4 hours, the catalyst was charged again with naphthalene-laden air under conditions identical to those before the shutdown, i.e. a salt bath temperature of 380° C., an air flow rate of 4 m.sup.3 (STP)/h, an o-xylene loading of 0 g/m.sup.3 (STP) and a naphthalene loading of 40 g/m.sup.3 (STP). After a further day, the product gas composition was analyzed (see Table 9).

(68) TABLE-US-00009 TABLE 9 Before the After the shutdown shutdown Days since shutdown 0 1 Air rate [m.sup.3 (STP)/h] 4.0 4.0 Loading [g.sub.o-X/m.sup.3 (STP)] 0 0 Loading [g.sub.naphthalene/m.sup.3 (STP)] 40 40 Salt bath temperature [° C.] 380 380 o-Xylene [% by wt.] 0 0 Phthalide [% by wt.] 0 0 Naphthoquinone [% by wt.] 1.267 0.516

(69) It has been found that, under identical conditions, the product quality after the interruption of the supply of the aromatic hydrocarbon to the catalyst has been improved, noticeable by the reduction in the content of naphthoquinone, an underoxidation product, with equal naphthalene concentration in the product.