Lloyd L. Handbook of Industrial Catalysts

Подождите немного. Документ загружается.

278 Chapter 7

ferent catalysts is used and high selectivity is achieved at fairly low conversion.

The Phillips Star process used a promoted noble metal catalyst supported on

zinc aluminate, which had been originally developed for light paraffin dehydro-

genation with minimum isomerization.

The catalyst is contained in a tubular

reformer, operating at temperatures in the range 480°–620°C and at a pressure in

the range 1–2 bar. The space velocity ranges between 0.5–1.0 LHSV and the

steam ratio is in the range 2–10 moles of steam per mole of hydrocarbon. The

process suffers from the deposition of carbon, and the catalyst must be regener-

ated after an eight hour cycle. The average lifetime of the catalyst is about 1–2

years. The selectivity to propylene is around 80–90% at a propane conversion of

30–40%. In the case of butane, the selectivity to butene is 85–95% at a conver-

sion of 45–50%. The UOP Oleflex process uses a platinum catalyst supported on

alumina under similar conditions and selectivity. A chromium-based catalyst is

used in the Catofin process. A more recent process, announced by Linde in 1992

and using a chromium/alumina catalyst developed by Engelhard, is now operat-

ing in Germany.

The platinum catalyst, promoted with tin and supported on

CuO/ZnO/Al

(derived from calcination of a mixed precursor with a hy-

drotalcite structure) is different from the early formulations. It is reduced in hy-

drogen and steam before use. High conversions and selectivities are claimed in

operation at high temperatures with extended cycle times.

7.2.4. Styrene

The Naugatuck Chemical Company, Connecticut, a subsidiary of U.S. Rubber,

was the first company to manufacture styrene. The process, however, was based

on the hydrodechlorination of ethylbenzene and the product was not sufficiently

pure for use in polymer applications. The company had received assistance from

Igor Ostromisslenski, a well-known Russian emigrant, who held several patents

in the field.

Both I. G. Farben in Germany

and the Dow Chemical Company, Midland,

Michigan,

were also working, independently, on styrene and polystyrene and

had applied for patents. By 1931 I. G. Farben were operating a 60-tonnes.day

−1

plant at Ludwigshafen and, soon afterward, Dow built a plant in the United

States. Ethylbenzene was dehydrogenated directly to styrene in both processes,

although different reactor designs were used. The catalysts used by I. G. Far-

ben

and Dow

were quite different. At first neither used potash, which is now

known limit the degree of carbon deposition. I. G. Farben used supported zinc

oxide whereas Dow used bauxite. The iron oxide impurity in the bauxite, as in

the Phillips butane dehydrogenation catalyst, would have acted as a fortuitous

catalyst promoter and is still an essential component of modern catalysts. Cata-

lyst compositions are shown in Table 7.5.

Early patent literature does not always give a true idea of the process. Much

technical information on various petrochemical processes became available in

Petrochemical Catalysts 279

the years following World War II, when experts inspected the German chemical

engineering and technical centers producing chemicals. The results were pub-

lished in a number of reports:

• British teams wrote British Intelligence Objectives Sub-Committee Re-

ports (BIOS).

• American teams prepared Field Information Agency Technical Reports

(FIAT).

• There were also Combined Intelligence Objectives Sub-Committee Re-

ports (CIOS).

The German process used banks of gas-fired, copper-lined tubes, 30 ft long

and 8 in in diameter, containing catalyst. These were heated to 610°C. The feed

to the reactor tubes consisted of a mixture of 1.25 lb steam for every pound of

ethylbenzene, preheated to 560°C. Up to 1941 the yield was 75% at 40% con-

version, but later the yield improved to 90%, presumably as a result of using an

improved catalyst. The US process used a fixed bed of catalyst with 2.6-lb steam

for every pound of ethylbenzene. Operation at 600°C gave a catalyst life of one

year with no regeneration. As with other dehydrogenation reactions, the partial

pressure of ethylbenzene was reduced by the addition of steam or even carbon

dioxide. Modern processes are based on the use of isothermal tubular reactors or

fixed bed adiabatic reactors. Unconverted ethylbenzene is recycled after the sty-

rene product is removed. When the Rubber Reserve Company authorized buta-

diene production in 1942, they also included six styrene plants, five of which

used the Dow process. The sixth plant used a Union Carbide process. A total of

187,500 tonnes/year of styrene was planned, based on ethylene and coal-based

benzene. When Standard Oil introduced catalyst 1707 for butadiene production,

it was also used to dehydrogenate ethylbenzene in the new plants. The improved

iron oxide/chromia/ potash catalysts and Shell 105 and 205 were eventually pre-

ferred when they became available, particularly after the war.

Many other iron catalyst compositions have been examined experimentally

as the styrene process has developed.

Chromium was replaced or combined

with a wide range of other oxide promoters, such as copper oxide, zinc oxide,

vanadium pentoxide, thoria, tungsten oxide, molybdena, ceria, and alkaline earth

oxides. At the time, none was used in any full-scale process but some were later

found to give good performances in improved plants.

7.2.4.1. Ethylbenzene Production

The alkylation process is the addition of an alkene to benzene, usually over an

acidic catalyst to give the alkyl benzene. The reaction is non-selective, and poly-

alkyl benzenes are regular impurities in the crude product stream. The degree of

polysubstitution is usually limited by controlling the ratio of reactants.

280 Chapter 7

TABLE 7.5. Ethylbenzene Dehydrogenation Processes.

Operating conditions

Catalyst volume (m

.100,000 tonnes

-1

year

−1

)

~40

Inlet temperature (

580–600

Outlet temperature (

540–600

Pressure (atm) <1–2

Liquid space velocity (LHSV) 0.3–0.7

Steam ratio (weight) 1–2

Steam ratio (molar) 6–12

Conversion (%) ~60

Selectivity (%) 93–96

Catalyst life (years)

Adiabatic

Isothermal

1–2

4–5

Catalyst

a,b

(a) (b)

%Fe

60 70–80

O 30 10

%Cr

1–2 —

%Ce

%MoO

2–3

%CaO

%MgO 7–8 2

%Loss at 540

C Balance Balance

Particle size (mm) 3–7 3–7

Bulk density (kg l

−1

)

1.0–1.8 1.0–1.3

Surface area (m

−1

)

2–3 2–3

Pore volume (cm

−1

) 0.2–0.3 0.2–0.3

Some catalysts contain binders (e.g., cement)

Early World War II catalysts compositions were: I.G. Farben (a) 72% ZnO; 18% Al

; 9%

CaCrO

; (b) 77.4% ZnO; 9.4% Al

; 2.8% CaO; 2.8% K

CrO

; 2.8% K

. Dow: bauxite (12%

) + 5% Ba(OH)

or KOH.

Dry benzene was alkylated with ethylene in either the liquid or gas phases

using acidic catalysts:

• An early liquid phase process used an aluminum trichloride catalyst at

85°– 95°C at pressures just above atmospheric. A low ethylene/benzene

ratio was used to limit the formation of diethylbenzene and other polyeth-

ylbenzenes. By-products could, however, be recycled with benzene and

were recovered as ethylbenzene by transalkylation. Ethylbenzene selec-

tivity was about 94% based on benzene and higher on ethylene. The cata-

lyst that formed in solution was thought to be HAlCl

.n-C

, which

gradually deactivated and was replenished as required. Other acid cata-

lysts such as boron trifluoride can be used in the liquid phase process,

which is still widely used in older plants.

Petrochemical Catalysts 281

• Gas phase processes have been used and operate at 300°C and higher

pressures up to 60 atm. The first catalysts were the typical phosphoric ac-

id/kieselguhr types (UOP Alkar process) or silica/alumina (Koppers) op-

erating with a low ethylene/benzene molar ratio (0.2). The low ratio was

necessary to minimize by-products that could not be dealkylated by the

catalyst. The Alkar process using boron trifluoride supported on activated

alumina did, however, dealkylate the by-products, thus giving higher se-

lectivity at 100% ethylene conversion. It was useful for the production of

ethylbenzene from dilute ethylene streams.

• Since 1980 high-activity modified ZSM-5 catalysts that operate at 450°C

and 20 atm have been used in most new plants. Capacities of up to

800,000 tonnes/year have been possible. The catalyst is extremely active

and forms few by-products or coke-forming polymers. Capital costs for

equipment are lower and the use of highly corrosive catalysts is avoided.

7.2.4.2. Styrene Production after 1950

Although the use of dehydrogenation processes to supply butadiene declined as

the more economical supplies from steam cracking of naphtha were introduced,

the production of styrene from ethylbenzene dehydrogenation has been continu-

ously developed, since styrene is not available in sufficient quantities as a by-

product.

From about 1950, Shell 205 and similar catalysts based on alkalized iron

and chromium oxides were used exclusively for styrene production. As plant

capacities were rapidly expanded, efforts were increased to improve the perfor-

mance of the catalyst. Higher potash levels were introduced

and cement bind-

ers were used to increase strength and selectivity.

Ethylbenzene conversion,

which was still about 30–50% in the 1950’s, was increased to at least 60% by

1960. Better plant designs were developed and reactors with up to three beds

were introduced. One of the first higher selectivity catalysts included vanadium

pentoxide with the conventional chromium oxide and potash.

Improvements

often led to different catalysts being used in a single reactor to optimize opera-

tion.

The energy crisis of the 1970s made it even more important to reduce costs

and make the process more efficient. Selectivity was increased even further by a

new catalyst, in which a combination of molybdenum oxide and cerium oxide

together with a cement binder replaced the chromium oxide completely. It was

still necessary to reduce fuel costs and operators began to conserve energy by

lowering the steam ratio during operation. Under these conditions chromium

promoters were still effective, and as the steam ratio decreased, chromium-

promoted catalysts, particularly those with high potash content and cement

binders, operated with a much lower level of deactivation for longer periods

than the new chrome-free catalysts. High-potash chromia catalysts operated at a

282 Chapter 7

steam mole ratio as low as 3:1, while catalysts containing mixed chromium,

ceria, and molybdenum promoters, together with potash, were more selective

but had to be operated at steam ratios of about 7:1.

Increased efforts during the 1970s led to a rapid increase in the availability

of catalysts containing various promoters and different potash contents. One

supplier estimated that some 12–15 catalysts were available for both isothermal

and adiabatic reactors operating at both high and low steam ratios.

This was

partly because so many different processes were used and every operator proba-

bly had a favorite type of catalyst! It was still possible, however, to choose cata-

lysts for both isothermal and adiabatic plants that had either a high selectivity

and good activity at a moderate steam ratio or a high activity and long life at a

higher steam ratio.

7.2.4.3. Styrene Plant Operation

The dehydrogenation of ethyl benzene is endothermic so that heat must be sup-

plied during operation. The two commercial styrene processes either incorporate

several adiabatic beds with interbed heat exchange/steam addition or isothermal

tubular reactors with a suitable heating medium in order to maintain operating

temperature.

Carbon deposition and dealkylation reactions are both inhibited by the addi-

tion of potash to the catalyst. The equilibrium of the forward reaction is favored

by high temperature and low operating pressures. A high proportion of super-

heated steam is also added to the ethyl benzene feed for the following reasons:

• Provide heat for the process.

• Reduce the partial pressure of ethylbenzene.

• Inhibit the formation of undesirable carbon deposits.

• Prevent reduction of magnetite to metallic iron.

After charging to the reactor, the catalyst is carefully heated, first in nitro-

gen and then in steam, to about 540°C, before ethylbenzene is admitted to the

reactor. A high steam ratio is maintained for several days to activate the catalyst.

During the activation period, the α-haematite (α-Fe2O3) component of the cata-

lyst is reduced to magnetite (Fe3O4) and selectivity increases to a maximum

during a period of two to three days. The level of conversion is maintained by

controlling the temperature of the catalyst or by increasing the steam ratio. The

presence of chloride in the feedstock or in the steam leads to an irreversible poi-

soning of the catalyst. Sulfur compounds also act as poisons, but the activity of

the catalyst does recover when the sulfur contamination is removed from the

feedstock. Operating conditions are summarized in Table 7.5.

Some catalysts can catalyze the dealkylation of the ethyl benzene to eth-

ylene and benzene, as well as the required dehydrogenation reaction to give sty-

rene and hydrogen. This undesirable reaction can be suppressed by allowing a

Petrochemical Catalysts 283

small amount of carbon or coke to deposit on the catalyst surface. However, an

excessive level of carbon deposition also inhibits the required dehydrogenation

reaction. When this occurs, some of the carbon can be removed, and activity for

hydrogenation restored, by treating the catalyst with superheated steam. If the

conditions under which the catalyst is steamed lead to conversion of the potassi-

um carbonate to the hydroxide, the catalyst selectivity decreases and potassium

hydroxide, which is volatile, is slowly lost. Catalyst life is usually less than two

years.

7.2.4.4. Ethylbenzene Dehydrogenation (Styrene) Catalysts

Catalysts require a high thermal stability to maintain the required surface area

and pore size at the high operating temperature and steam ratio.

The composi-

tion of the catalyst is quite similar to the high-temperature-shift catalyst used in

the manufacturing of ammonia, which is operated at much lower temperatures

and which often loses strength when exposed to potassium compounds which

have migrated from the reforming stages of the process. Perhaps for this reason,

hydraulic cements have often been used to bind styrene catalysts and improve

physical strength. Catalysts have been made by wet mixing of the components to

form a paste.

Early catalysts were produced from calcined ferric oxide, potassium car-

bonate, a binder when required, and usually chromium oxide. Subsequently a

wide range of other oxides replaced the chromium oxide;

51,52

typical composi-

tions are shown in Table 7.5. The paste was extruded or granulated to produce a

suitable shape and then calcined at a high temperature in the range 900°–950°C.

Solid solutions of

-hematite and chromium oxide (the active catalyst precur-

sors) were formed and these also contained potassium carbonate to inhibit coke

formation. Catalyst surface area and pore volume were controlled by calcination

conditions. It has been confirmed by X-ray diffraction studies that α-hematite is

reduced to magnetite and that there is some combination of potash and the

chromium oxide stabilizer. There is little change in the physical properties of the

catalyst during reduction and subsequent operation.

Catalyst activity and selectivity, at high or low steam ratios, can be con-

trolled by selection of both the stabilizers and the calcination temperature.

7.3. SYNTHETIC FIBERS

The production of nylons and polyesters, which were first synthesized in the

1930s, has led to the development of man-made fibers. The complex sequence

of catalytic reactions involved in both processes was originally based on 1930s

technology but, as the reaction mechanisms came to be understood, catalysts

were improved and, in some cases, new petrochemical feeds were introduced.

284 Chapter 7

Acrylic fibers based on acrylonitrile were then developed from about the 1940’s

and, although not as widely used as polyesters and nylons, have many applica-

tions.

It is shown in Table 7.6 how synthetic man-made fibers are replacing the

viscose and acetate rayons produced from natural cotton fibers or wood pulp.

During 1970 the production of cellulose fibers was 43% of world fiber output,

but by 2000 the proportion had fallen to only 8%. Cellulose fiber production fell

from 3.6 to less than 2.5 million tonnes during the same period. Polyester fiber

dominates the market and far exceeds the use of nylon 66, nylon 6, and acrylic

fibers. In 1993 man-made synthetic fiber production was about 17.5 million

tonnes while man-made cellulose fiber production was about 2.5 million tonnes.

At the same time about 15.5 million tonnes of natural cotton was used.

7.3.1. Nylon 66

Wallace Carothers began his work on polymerization in 1928 at the DuPont

laboratories. By 1930, he had found that aliphatic polyesters, formed by the

condensation of dihydric alcohols and dicarboxylic acids, could be drawn into

threads when melted.

Most of the polyesters he investigated, however, had low melting points and

were too soluble in water for use as textiles, although polyamides were more

promising. To form useful fibers, the polymer needs to have a molecular weight

in the range12,000–20,000. When amino acids were briefly investigated, those

with fewer than five carbon atoms condensed to form lactams, together with a

low-molecular-weight polymer. Carothers then assumed that only those amino

acids containing more than seven carbon atoms would be able to produce useful

polymers. These were difficult to synthesize, so he concentrated on the range of

polyamides produced from diamines and dicarboxylic acids. He selected adipic

acid and hexamethylene diamines for his work because cyclohexane was readily

available and could be used as the raw material for conversion to both interme-

diates. The polymer produced from the two C

molecules formed good fibers

TABLE 7.6. Production of Synthetic Fibers 1970–2000.

Fiber 1970 1993 2000

% Polyester 20 50–60 62

% Nylon

24 15–20 12

% Acrylics 12 ~10 9

% Cellulose 43 ~10 8

% Others

1 ~ 5 9

Total (million tonnes) 8.5 20 30

Nylon also used, e.g., for carpets, and polyester also used, e.g., for bottles.

Polyethylene, polypropylene, polyvinyl alcohol, polyvinyl chloride.

Petrochemical Catalysts 285

and production of nylon 66 began in 1938.

The term nylon 66 is derived from

the fact that the diamine and the dicarboxylic acid both contain 6 carbon atoms.

7.3.1.1. Production of Nylon Intermediates

During the early years of nylon 66 production, cyclohexane was obtained by

distillation from some natural gasolines, together with isomerization of the small

volumes of methyl cyclopentane also present. In other parts of the world and as

the demand for nylon fibers and resins increased, benzene became the major

source of cyclohexane.

Sulfur-free benzene can be converted to cyclohexane by hydrogenation in a

two stage process. Approximately 95% of the benzene in a circulating liquid

phase is converted in the first reactor using a simple Raney nickel catalyst at

temperatures less than 230°C and at pressures up to 50 bar. The catalyst also

acts as a sulfur guard. Temperature control is achieved largely by relying on the

latent heat of evaporation of cyclohexane. It is important to avoid isomerization

to methyl cyclopentane. A second reactor with a fixed catalyst bed of nickel

supported on kieselguhr or alumina, operating in the vapor phase at 200°–225°C

and 50 bar, completes the benzene conversion.

7.3.1.2. Adipic Acid

The first stage in the production of adipic acid is the oxidation of liquid cyclo-

hexane with air using a cobalt naphthenate catalyst at temperatures in the range

at 140°–160°C and pressures about 8–12 bar. A mixed ketone/alcohol oil con-

taining cyclohexanol (C

OH) and cyclohexanone (C

O) (KA oil) is pro-

duced at up to 85% selectivity, with the ketone/alcohol ratio of about 1:1. The

conversion is only about 10% and unconverted cyclohexane is recycled. The

process was improved by Bashkirov by the additions up to 5% boric acid to the

cyclohexane and by restricting the oxygen content of the air to about 4%. The

overall yield was increased to more than 90% and conversion to more than 12%.

The ketone/alcohol ratio was decreased to 1:9.

The ketone/alcohol oil was then oxidized further to adipic acid in a 60%

nitric acid solution with an ammonium metavanadate/copper nitrate catalyst at

50°–80°C. Selectivity to adipic acid was more than 95%. A second air oxidation

process at 80°–85°C and 6 bar has since been developed. By using a copper ace-

tate/manganese acetate catalyst in acetic acid solution, the process has the ad-

vantage of avoiding the use of the strongly corrosive nitric acid:

OH + 4 O(HNO

) → COOH(CH

)

COOH + H

O (7.26)

O + 3 O(HNO

) → COOH(CH

)

COOH (7.27)

286 Chapter 7

Pure cyclohexanol can also be obtained by the hydrogenation of phenol

with a palladium catalyst at 150°C and 10 atm although the process is not widely

used. Attempts were made by BASF to synthesize adipic acid from butadiene

via a two-step carbonylation process in the presence of methanol. The first step

of the synthesis operated at 130°C and 600 bar while the second operated at

170°C and a lower pressure of 160 bar. A typical cobalt catalyst with organic

ligands was used, but the process was never developed industrially.

Phenol produced by the cumene–phenol process is relatively expensive but

Solutia has recently claimed a new process developed in Russia.

Benzene is

oxidized directly to phenol using nitrous oxide. Phenol is then converted to adip-

ic acid by oxidature procedures. Nitrous oxide from the final nitric acid oxida-

tion to adipic acid is recycled to the first stage. This has been reported as the

first new commercial technology since DuPont introduced the direct hydrocya-

nation of butadiene to adiponitrile in 1971.

7.3.1.3. Hexamethylenediamine

The manufacture of hexamethylenediamine is usually a two stage process. In the

first stage, adiponitrile is produced, and in the second stage, adiponitrile is hy-

drogenated to hexamethylenediamine. Adiponitrile has been manufactured, in

the main, from three feedstocks, and these will be described briefly in turn.

In the first process adipic acid was converted to adiponitrile by a high tem-

perature reaction with ammonia over a boron phosphate catalyst at temperatures

in the range 300–350◦C. The process was thought to proceed via the formation

of diammonium adipate, followed by dehydration to adiponitrile:

COOH-(CH

)

-COOH + 2 NH

→ COONH

(CH

)

-COONH

(7.28)

COONH

(CH

)

-COONH

→ NC(CH

)

CN + 4 H

O (7.29)

The selectivity to adiponitrile was about 80% in the early reactors.

The main problem with this reaction is that adipic acid in the liquid phase is

thermally unstable and degrades to a carbon-rich, hard, resinous product. The

vapor is rather more thermally stable at a similar temperature, but at higher tem-

peratures also degrades to similar products. The overall reaction producing adi-

ponitrile from adipic acid is very endothermic, and the reaction mixture is sub-

jected to a significant measure of cooling due to this process. Thus, while all of

the reaction components were initially in the gas phase this cooling result in the

condensation of some liquid adipic acid, which then proceeds to degrade to the

polymeric resin described above. The reactor tubes regularly become blocked

with catalyst and polymeric resin, thus preventing flow of reactants through the

reactor. At this stage, the reactor must be taken out of service, so that the poly-

meric resins can be drilled out of the reactor, and the catalyst replenished. Selec-

Petrochemical Catalysts 287

tivity could be increased to 90% by converting the adipic acid to adiponitrile in a

fluid bed with a phosphoric acid/kieselguhr catalyst that was continuously re-

generated during operation.

As the demand for nylon increased during the 1950s and 1960s, an alterna-

tive, more readily available feedstock was sought for the production of adiponi-

trile. In this way more adipic acid would be released for the production of nylon

salt. Consequently, by the 1960’s, an alternative route to the production of adi-

ponitrile using butadiene as the raw material was sought, either by direct or indi-

rect hydrocyanation. The indirect route proceeded via chlorination with no cata-

lyst, at 200°–300°C, followed by the replacement of chlorine with cyanide to

produce a mixture of dicyanobutene isomers. The mixture was isomerized to the

required 1,4-dicyano-2-butene by a copper complex in a reaction mixture con-

taining hydrogen cyanide. Adiponitrile was formed by gas phase hydrogenation

at 300°C using a palladium catalyst.

The direct addition of chlorine to butadiene followed by reaction with sodi-

um cyanide is a very unattractive process. Both chlorine and sodium cyanide are

very expensive to produce in terms of energy required, and then to be left with

the problem of disposal of contaminated sodium chloride residues only serves to

compound the problem. The direct addition of hydrogen cyanide to butadiene is

much more attractive, provided that the addition reactions can be tailored to

produce the desired linear product. The problems were solved by some excellent

work by scientists from du Pont.

56,57

The chemistry of the process can be envis-

aged in three stages. In the first stage, one molecule of hydrogen cyanide adds

across one of the double bonds to give a mixture of linear and branched nitriles.

In the second stage, the branched isomer is isomerized to the linear form, and in

the third stage, the second molecule of hydrogen cyanide adds across the re-

maining double bond to give adiponitrile:

=CH.CH=CH

+ HCN → (7.30)

=CH.CH

CN + CH

=CH.CH(CN).CH

=CH.CH(CN)CH

→ CH

=CH.CH

CN (7.31)

=CH.CH

CN + HCN → NC(CH

)

CN (7.32)

The catalyst is similar for all three steps, and consists of a zero valent nickel

phosphite complex, promoted with zinc or aluminium chlorides. The direct addi-

tion of hydrogen cyanide to butadiene is particularly attractive with the availa-

bility of by-product hydrogen cyanide form the manufacture of acrylonitrile by

the ammoxidation of propylene.

The third of the potential feedstocks for the manufacture of adiponitrile is

acrylonitrile, prepared by the ammoxidation of propylene. The only route to