Lloyd L. Handbook of Industrial Catalysts

Подождите немного. Документ загружается.

418 Chapter 10

gas, and hence a significant difference to the economics of the process. Any real

improvement to the process was therefore dependent upon the development of a

new catalyst with significantly improved activity. The clear target would be a

catalyst with sufficient activity to give a satisfactory level of conversion at the

same pressure as that within the reformer.

10.1.5.1 Magnetite Catalyst Containing Cobalt

The surface area of the ammonia synthesis catalyst is only about 1–2 m

-1

. It

was known from experience with other catalysts that precipitation from aqueous

solution always led to a product with a much higher surface area than one pre-

pared by fusion, and therefore potentially at least, more active sites per unit

weight of catalyst. This process route was studied in ICI, initially by Topham.

Subsequent developments led to the preparation and testing of a precipitated

catalyst containing cobalt which showed levels of activity approximately 3 fold

higher that of the best conventional catalyst available at the time.

Ammonia

could be synthesised at temperatures below 350ºC, but the catalyst was not

commercialised for several reasons. The major reason was that the pelleted ox-

ide was significantly weakened during the reduction procedure, and became too

weak to withstand the rigours of an axial flow converter. It also suffered from

shrinkage during reduction, leading to settling of the bed, and the likely devel-

opment of hot-spots.

The ICI cobalt catalyst, which is reported to contain 10-20% cobalt oxide,

was developed for use at 80 bar synthesis pressure and maximum temperatures

as low as 460°C. These conditions were significantly different from the usual

single-stream ammonia plants which operated at 150 bar or higher and tempera-

tures up to 500°C.

From experimental work with precipitated cobalt-iron catalysts it appeared

that the cobalt reacted with alumina giving cobalt spinels, which helped to form

smaller iron and cobalt crystallites during reduction and to increase the surface

area to make the catalyst more stable and less sensitive to oxide forming poi-

sons. In the presence of potash, cobalt was able to decrease the reduction tem-

perature.

Development work was successful and, following a typical fusion produc-

tion procedure, the new catalyst could be reduced and operated at lower temper-

ature under normal conditions in conventional plants. Reduction began at 380°C

in the new AMV and LCA processes introduced by ICI in 1985 and 1988 re-

spectively. The AMV plants have operated at more than 100% capacity of a

pressure of only 60 bar. The life of the new catalyst has been very satisfactory,

and could be operated of pressures much closer to those of the reformers, albeit

with large converters.

Ammonia and Methanol Synthesis

419

10.1.5.2 Ruthenium Catalyst

The announcement in 1979 by BP of a ruthenium catalyst was the first real ad-

vance in improving the process. The catalyst was more active and operated at a

lower temperature to produce a higher equilibrium ammonia concentration. A

relatively high operating pressure was still needed, however, when using a bed

charged with Ru catalyst in conjunction with other beds containing magnetite

catalysts. The catalyst was reversibly poisoned by some of the impurities pre-

sent in typical synthesis gas.

10.1.5.3 Catalyst Preparation

Little information is available on the commercial production of the ruthenium

synthesis catalyst. A typical catalyst prepared according to a Canadian patent

has, however, been compared with typical iron catalysts. Many other patent ref-

erences have also been included in a review of non-iron catalysts. Descriptions

stress the importance of the thermal treatment of the carbon support to control

its structure.

Carbon is heated up to at least 1000°C in an inert or reducing atmosphere.

The acid surface of the carbon then becomes basic with no effect on the pore

structure. The heat treatment deactivates the carbon against the reaction with

hydrogen to form methane during subsequent operation. Commercial coconut

charcoals, or those produced from hard woods or cellulose, can be used. It is

important to use a support with a relatively low surface area of less than 700

/g.

The carbon can be impregnated with the active metal and two promoters, in

the following sequence, being dried and heated to 250°C between each stage:

• 2% of barium as nitrate;

• 4% of ruthenium as chloride which is more stable than other salts and

gives a better metal distribution;

• 12% of potassium or cesium as hydroxide.

A black, shiny catalyst is produced. Vacuum impregnation can give a uni-

form distribution of the metals in the carbon although this procedure may not be

Two important points are significant with catalysts prepared in this way.

The order of impregnation influences the catalyst activity and the catalyst is

used for large-scale production. Addition of ruthenium as a volatile carbonyl was

also considered during early development. Before use, the catalyst is activated

been removed.

by heating in synthesis gas for twelve hours, at a temperature of 400 C

and pressures up to 25 bar, until more than 95% of the chloride content

has

420 Chapter 10

sensitive to hydrogen concentration. A typical 3:1 hydrogen-nitrogen ratio inhib-

its the activity of ruthenium. Much better operation is possible at temperatures

as low as 400°C with 1:1 hydrogen-nitrogen ratio.

Further work has shown that cesium is a better promoter than potassium and

that the alkali hydroxides are, surprisingly, reduced to a metallic state during

catalyst activation. This seems thermodynamically improbable, but may result

from the high heat of adsorption of the alkali metal hydroxides on carbon, which

leads to charge transfer to the carbon, and could drive the reduction of the hy-

droxide to the metal. The alkali promoter may neutralize residual chloride ions

and develop catalyst activity during reduction as ruthenium ions diffuse from the

lattice to form crystallites on the edges of graphite crystals.

Ruthenium, promoted by alkali metals and barium and then supported on

carbon, has provided a significant increase in activity compared with iron cata-

lysts. The high cost of ruthenium when compared with iron must have led to it

being used with the cheap support material. In retrospect, the complex interac-

tion between carbon, alkali metal and barium with the active ruthenium led to

the development of a successful catalyst. A magnesium oxide support has now

been claimed to give a longer life than carbon.

10.1.5.4 Full-scale Operation with Ruthenium Catalyst

The new ruthenium catalyst has been used in the Pacific Ammonia Inc plant at

Kitimat in British Columbia since 1992. Ammonia had been made there, from

the methanol plant purge gas, since 1986 using a conventional iron synthesis

catalyst. The new ruthenium catalyst converter was in series with the old con-

verter. Although in 1992 there was no additional synthesis gas to increase pro-

duction capacity, the ruthenium catalyst operated well in a radial flow reactor

and reduced both the steam and electricity used by 30-40% and 5-10% respec-

tively. The new catalyst was said to be twenty times as active as the iron cata-

lyst, and the effluent gas contained about 20% ammonia.

The catalyst was reduced at 300ºC, and took less than one day, with the

evolution of only small amounts of water. The catalyst was operated at lower

pressures and temperatures than the iron catalyst, and in marked contrast, could

also tolerate the presence of 4000 ppm of carbon monoxide, with little or no

effect on activity.

Other large ammonia plants are now using single beds of the ruthenium cat-

alyst in conjunction with the magnetite catalyst. They confirm the higher activity

of ruthenium, with the benefits of a lower operating pressure and temperature,

while maintaining a high concentration of ammonia in the exit gas. For future

process designs, the need for less gas compression and less high-pressure

equipment will lead to lower operating and capital costs. Two plants in Trinidad,

using one bed of iron catalyst and three beds of ruthenium catalyst, have operat-

ed for several years.

It is reported that the capacity of typical 1100 tonnes per

Ammonia and Methanol Synthesis

421

day plant designs could be increased to 1850 tonnes per day, provided there is

sufficient reformer capacity.

10.2 METHANOL SYNTHESIS

Experiments during the nineteenth century had shown that certain oxides could

be used as dehydrogenation catalysts. For example, Jahn dehydrogenated meth-

anol by passing the vapour over finely divided zinc or zinc oxide to produce a

stoichiometric mixture of hydrogen and carbon monoxide.

Subsequently, patents covering the conversion of synthesis gas to complex

mixtures of organic oxygen compounds, including methanol, were issued to

BASF during 1913. This followed work by Mittasch and Schneider. Full-scale

production of methanol was not attempted, however, until 1923. By that time

high-pressure equipment had been in operation for several years in the new am-

monia process. The methanol process was developed by Piers and the plant,

built at Leuna, used mixed zinc oxide-chromic oxide catalyst. The use of metal-

lic iron for the internal parts of the reactor was avoided to prevent the formation

of the volatile iron pentacarbonyl. The would have decomposed on the surface

of the catalyst, to deposit finely divided iron metal, which in turn would have

promoted the exothermic formation of methane.

10.2.1 High-Pressure Synthesis

Although coal based water gas was used as feed to the methanol plant the cata-

lyst was probably developed in conjunction with work on coal hydrogenation

and the Fischer Tropsch process. The aim of the process had been the production

of a range of liquid hydrocarbons but it was found that the high-pressure reac-

tion with water gas formed very pure methanol. The methanol was tested as a

fuel for automobiles, but this was unsuccessful because of knocking in the low

compression engines then being used. Other applications for methanol were

soon developed as, for example, bakelite was introduced.

10.2.1.1 Zinc Oxide-Chromium Oxide Catalysts

The BASF catalysts were made by mixing solid chromic acid with an aqueous

suspension of zinc hydroxide, drying the paste and forming granules. Later an

even simpler catalyst was made by mixing zinc oxide powder with chromic ox-

ide, making a paste with water and then granulating. The catalysts were not

particularly active and heat evolution had to be carefully controlled during re-

duction.

During the 1920s the methanol process and various catalysts were widely

studied both in Europe and in the US. It has been assumed that Patart, in France,

422 Chapter 10

investigated zinc oxide catalysts, containing other metals, as a result of Jahn’s

observations. Patart operated a semi-technical scale reactor with a zinc oxide

catalyst supported on asbestos. The patent, issued in 1921,

and other published

work gave many details of methanol production. Later on Patart worked with

zinc oxide catalysts containing copper oxide although best results were obtained

with a 3:1 molar mixture of zinc oxide and chromic acid.

Active and stable catalysts containing an excess of zinc oxide were eventu-

ally made by co-precipitation of soluble zinc and chromium salts. Natta did,

however, show that the natural zinc carbonate, Smithsonite, which contained

impurities such as cadmium, magnesium and copper, had a high stable activity.

While copper oxide in combination with zinc oxide or other supports was an

active catalyst, it was unstable at high temperature and easily poisoned by the

impurities in water gas.

From about 1930 onwards, all catalysts used industrially, for example by

BASF, Du Pont, Montedison and ICI, were based on zinc oxide stabilized by the

high melting chromium oxide. The actual composition depended largely on

whether the catalyst was precipitated or was just a simple mixture of oxides.

The catalysts were prepared either by mixing or precipitation of the compo-

nents, drying, followed by calcination in the temperature range, 400–450ºC. The

catalyst powder was then compacted and formed into granules or pellets. When

the temperature of the catalyst rose above 200ºC, it was found that the chromi-

um oxide component, particularly in the case of precipitated catalysts, was par-

tially oxidised to the hexavalent state. However, as the temperature rose higher,

the hexavalent chromium decomposed back to the trivalent state. Consequently,

the high chrome catalysts prepared by mixing the oxides must have been partial-

ly pre-reduced before use, because of the redox cycles. Nevertheless, they often

cracked or lost particle strength. All precipitated catalysts had to be calcined at a

high enough temperature to limit the presence of hexavalent chromium, which

Zinc oxide-chromium oxide catalysts are often referred to as zinc chromite

but there is always a considerable excess of zinc oxide. The catalysts are rela-

tively stable at temperatures when zinc oxide alone would lose activity, the

chromium, perhaps present as a spinel,

is believed to stabilize the zinc oxide.

Maximum activity was claimed for precipitated catalysts containing 20-30%

chromium oxide.

However, these catalysts were rather unstable, due to shrink-

age in use, and longer life could be achieved with a lower chromium contents.

From the limited information available, precipitated industrial catalysts general-

ly contained less than 20% chromium oxide as shown in Table 10.8.

evolved heat during reduction in the reactor, leading to thermal processing and

the precipitation and calcination conditions.

sintering of the catalyst. Catalyst activity was shown to be proportional to the

surface area of the zinc oxide and that these properties could be controlled by

Ammonia and Methanol Synthesis

423

TABLE 10.8. High-Pressure Methanol Catalysts.

Composition Catalyst Type

Mixed Oxides Low Chromium High Chromium

Zinc oxide 59–64 69 80 67–68

Chromium oxide 30–26 (CrO

) 21 (CrO

) 11–13 20–21

Loss at 900°C

9–10 10 Balance 9–10

Typical impurities:

Sulfate

Sodium

Carbonate

Water

up to 1%

<0.5%

up to 10%

<3%-calcined 430°C

<4-5%-calcined 400°C

It was suggested that the relatively unstable mixed oxide catalysts could be

supplied in the pre-reduced form that was, effectively, pre-shrunk. While these

catalysts were offered for use commercially, there is no evidence that operation

was actually improved. In any case, the maximum operating temperature in a

high-pressure methanol plant was limited to 390°C by the onset of methanation

so that a stable activity was more important than high activity in the multibed

reactors used.

Lazier of Du Pont published many patents covering the preparation of metal

chromites. These were formed by precipitation from a solution of zinc nitrate

and chromic acid with ammonia at pH 6.8. The zinc ammine complex obtained

was decomposed at about 400°C to give the mixed oxides. As Adkins noted, his

copper chromite equivalent was extracted with dilute acetic acid solution to ad-

just the copper content. It is not clear whether the same treatment was ever used

in producing methanol catalysts or even whether the zinc ammine intermediate

was produced commercially. One problem with the Lazier preparation was the

difficulty in controlling the exothermic decomposition of the ammine that could

affect the catalyst activity.

Other companies preferred to precipitate the catalyst from solutions of the

cheaper zinc sulfate and chromium sulfate (chrome tan) with sodium carbonate.

An active catalyst, with a long life, could be produced although only about 90%

of the metals were recovered. This resulted from the variable pH of the solution

and the partial solubility of the precipitates at about pH 7 during the precipita-

tion procedure. Significant amounts of sulfate and sodium, in the form of chro-

mium complexes, also remained in the final catalyst although this did not appear

to result in loss of activity or to cause operating problems.

10.2.1.2 High-Pressure Operation

Most methanol plants operating before 1965 were small, producing about 150

tonnes per day at 250–260 bar pressure. The small catalyst volume of up to 4.5–

424 Chapter 10

TABLE 10.9. High-Pressure Methanol Process (150 Tonne Day

-1

, 260 Bar).

Variable Catalyst/Temperature Distribution

Bed number 1 2 3 4 5

Catalyst volume (m

) 1.2 0.6 0.8 0.9 1.0

Temperature inlet (°C)

340 360 360 360 360

Temperature outlet (°C)

364 375 380 380 375

Quench cooling between beds

Methanol exit reactor (vol%) ~3

Synthesis Gas Composition Volume %

Carbon monoxide 9

Hydrogen 78

Carbon dioxide 0.5

Methane 10

Nitrogen 2.5

5.0 cubic meters was divided into five beds, with inter-bed cooling to limit the

temperature rise, in a one-meter diameter reactor. The recycle loop, similar to

that in ammonia synthesis, had a high circulation rate of 150,000 m

-1

so that

space velocity in the catalyst was up to 30,000 hr

–1

. A typical set of operating

conditions is shown in Table 10.9 for a plant of this kind, which also shows the

temperature increase in each bed, the methanol concentration in effluent leaving

the final bed and the composition of synthesis gas entering the reactor.

About 2–5wt% of organic byproducts including about 1% of dimethyl ether,

apart from water, formed during the reaction and the amount usually increased

as the catalyst aged (a list of by-product formation is shown in Table 10.14.) A

significant amount of methane was also formed, particularly at temperatures in

excess of 390ºC. Much of this was formed on metallic iron, deposited on the

catalyst by the thermal decomposition of iron pentacarbonyl. Thus, the maxi-

mum operating temperature of the catalyst was severely limited. And the tem-

perature of the reactor was controlled by inter-bed quench cooling with cold

synthesis gas.

The loss of activity and selectivity from catalyst sintering and deactivation

increasing the bed temperatures. The main operating problem was usually an

increase in pressure drop. For this reason, after about two years’ use, the catalyst

was carefully removed from the beds in layers and sieved to remove dust. About

70% of the catalyst could often be replaced providing that it had adequate physi-

cal strength and an acceptable surface area, exceeding about 35 m

/gm. A typi-

cal catalyst life, after two or three sieving cycles, was 4–5 years. Some surface

area measurements on discharged samples from a five-bed unit are given in Ta-

ble 10.10. The surface area of new, precipitated catalyst was about 50m

/gm

although, following reduction, the surface area increased to 65–70m /gm.

from the deposition of iron and solids in the bed were partly compensated by

Ammonia and Methanol Synthesis

425

TABLE 10.10. Surface Area of Samples Used for Two Years.

Position Surface Area (m

-1

)

Bed 1 2 3 4 5

Top of bed 65 50 55 45 50

Middle 50 50 35 40 40

Bottom 40 50 40 35 35

Re-use catalyst >40 m

-1

Recovery up to 70% of volume

Overall life 4–5 years

mixed oxides— had to be carefully reduced in process gas to convert hexavalent

chromic oxide to trivalent chromium oxide.

10.2.2 Low-pressure Synthesis

The first industrial methanol catalysts, used from 1923, were based on zinc ox-

ide-chromium oxide mixtures. Experimental work during development of the

process had, however, demonstrated the high activity of zinc oxide catalysts

containing copper. This was particularly true if a third refractory oxide such as

chromia or alumina was also added.

Catalysts containing copper are particular-

ly susceptible to poisoning. At that time the synthesis gas was produced from

water gas, and contained poisons such as sulfur and chlorine compounds so that

copper catalysts were unsuitable.

When the hydrocarbon steam reforming process could provide poison free

synthesis gas, the benefits of the more active copper catalysts were quickly re-

viewed. It was soon shown that the high activity of copper oxide-zinc oxide cat-

alysts, compared with the zinc oxide-chromium oxide types, particularly when

alumina or chromia promoters were added, could revolutionize methanol pro-

duction. New, more efficient processes were urgently needed in view of the in-

creasing demand for methanol and the economies of high capacity units.

TABLE 10.11. Micromeritics of High Pressure Methanol Catalysts.

Measurement

Catalyst Type

Mixed Oxide Mixed Oxide

(reduced)

Low Cr High Cr

Surface area (m

-1

) 1.6 17.5 51 51

Pore volume (mlg

-1

) 0.04 0.23 0.38 0.17

Mean pore radius (nm) 45 26 13 6.5

Table 10.11. Before their use, the catalysts—particularly those prepared from

Micromeritic measurements on the catalysts shown in Table 10.8 are given in

426 Chapter 10

TABLE 10.12. Low Pressure Methanol Catalysts.

Licensor Commercial Catalysts Reference

ICI CuO.ZnO.Al

DOS 2302658 London (1973)

Lurgi CuO.ZnO.Cr

German Pat 1300917 (1969)

CCI CuO.ZnO.Al

DOS 1956007 Louisville (1969)

Mitsubishi CuO.ZnO.Cr

DOS 2165378 Tokyo (1971)

Topsoe CuO.ZnO.Cr

BASF CuO.ZnO.Al

CuO.ZnO.MnO

.Cr

CuO.ZnO.MnO

.Al

.Cr

CuO.ZnO.MnO

.Al

DOS 2056612 Frankfurt (1970)

DOS 1930003 (1969)

DOS 2026165 (1970)

DOS 2026182 (1970)

10.2.2.1 Copper Oxide Catalysts

The first full-scale trial of a copper catalyst to be reported took place in the

Polish Chemical works at Oswiecim during 1963 but was not successful.

Tests

showed that the precipitated copper oxide-zinc oxide catalysts available at the

time were unstable and either sintered or were poisoned very quickly. Better

formulations were investigated to develop large-scale production procedures and

increase catalyst stability. Improved catalyst testing procedures provided de-

tailed information on the catalyst structure and activity. This demonstrated that

the unstable catalysts contained large oxide crystals and that the metals were

distributed unevenly within the structure. Better catalysts required very small

crystals which were uniformly distributed.

It was soon found that catalysts

containing copper oxide-zinc oxide-alumina could be made with the appropriate

properties and a large, new plant was operating successfully by 1966.

Only one

new high-pressure methanol plant has been built since the low-pressure process

was introduced.

Soon after the introduction of the catalyst stabilized with alumina, others,

containing chromia, were also being produced. Processes based on both of the

new catalysts were operating from 1971

and many patents were published as

shown in Table 10.12.

Synthesis gas was obtained from a variety of sources including the steam

reforming of natural gas or naphtha and oil gasification. A flowsheet for a typi-

cal low-pressure methanol synthesis plant is given in Figure 10.8. The different

reactor designs used in methanol synthesis are listed in Table 10.13 and some

are shown in Figures 10.9 and 10.10. The correct carbon dioxide, as well as car-

bon monoxide, for the synthesis reaction to proceed.

10.2.2.2 Copper Catalyst Production

Until 1960, copper oxide-zinc oxide catalysts were usually precipitated in batch-

es. As with the zinc oxide-chromium oxide catalysts, an alkaline solution, such

as sodium carbonate, was added to the relatively acidic solution of the metal

Figur

natura

sion o

10.8. Flow sheet o

gas. Reprinted fro

M. Twigg.

typical low-

ressu

Catalyst Handboo

e methanol synthes

, 2

ed., Ed. by M.

s plant based on th

wigg, Copyright (1

steam reforming o

989) by kind permis

Ammoniia and Methanol Synthesis

4427