AU595967B2 - Cobalt-ruthenium catalysts for fischer-tropsch synthesis - Google Patents

Description

Api~ai~ COMMONWEALTH OF AUSTRALIA PATENTS ACT 1952-69 COMPLETE

SPECIFICATION

Form Number: Lodged:

(ORIGINAL)

Class I t. Class ~996 7 t'Gomplete Specification Lodged: Accepted: Published: Related Art Name Applicant: Address of Applicar~t Actual Inventor: Address for Service EXXON RESEARCH AND ENGINEERING

COMPANY

P.O. Box 390, Florham Park, New Jersey 07932, United States of America ENRIQUE tGLESI ROCCO ANTHONY FIATO and STUART LEON

SOLED

EDWD. WATERS SONS, 50 QUEEN STREET, ME~LBOURNE, AUSTRALIA, 3000.

Complete Specification for the Invention entitled: COBALT-RUTHENIUM CATALYSTS FOR F ISCHER-TROPSCH SYNTRE8SS The following statement Is a full dlescripthn of this Invention, Including the best method of performing It known to s i i i -1- COBALT-RUTHENIUM CATALYSTS FOR FISCHER-TROPSCH SYNTHESIS BACKGROUND OF THE INVENTION Field of the Invention This invention relates to an improved catalyst for producing hydrocarbons from synthesis gas, hydrogen and carbon monoxide, and to improvements in the hydrocarbon synthesis process. Specifically, this invention relates to a catalyst comprising cobalt and ruthenium in catalytically active amounts on a titania support and a process for utilizing the catalyst that allows on-stream regeneration and cyclical operation without having to remove the catalyst from the hydrocarbon synthesis reactor.

The Prior Art I Methane is available in large quantities in many areas of the world. Some methane is generated From refinery applications while large amounts of methane, as the principal constituent of natural gas, are found in deposits in various areas, Methane can 4 be used as a gas, for example, for heating purposes, arid can be transported by pipeline or as a liquefied gas over long distances. Where use of the methane as a gas is not economic or the transportation of methane requires traversing oceans, the methane can be converted to a liquid which is more easily transported and may have significantly higher value than methaie gas.

i ItI -2 Conversion of methane is normally carried out in a two-step procedure involving reforming the methane to produce hydrogen and carbon 3 noxide, synthesis gas, and converting the synthesis gas to higher hydrocarbons, C 5 in a Fischer-Tropsch type reaction. Both steps of the process are well known and can be readily illustrated: the first step by U.S.

Patents 1,711,036, 1,960,912 and 3,138,438; the second step by U.S. Patents 4,477,595, 4,542,122, and 4,088,671.

This invention is concerned with the second a step, the well kt own Fischer-Tropsch type reaction which will be referred to hereinafter as hydrocarbon

S

t synthesis.

This invention is primarily concerned with cobalt and ruthenium catalysts for hydrocarbon synthesis and both of these metals have been disclosed as being useful in such reactions, either alone, jointly, or with other materials. What has not been disclosed in the art is the combination of steps required to produce a composition that is novel and has superior catalytic properties to other cobalt, ruthenium, or cobalt-ruthenium catalysts. These c properties include: improved CO conversion, improved volumetric productivity, enhanced selectivity to and lower CH 4 and the ability to regenerate the catalyst at relatively lcw temperatures without removing it from the reactor.

U.S. Patent 4,477,595 discloses ruthenium on titania as a hydrocarbon synthesis catalyst for the production of C 5 to C 4 0 hydrocarbons with a majority

I

3 of paraffins in the C 5 to C 2 0 range. U.S. Patent 4,542,122 discloses a cobalt or cobalt-thoria on titania having a preferred ratio of rutile to anatase, as a hydrocarbon synthesis catalyst. U.S. Patent 4,088,671 discloses a cobalt-ruthenium catalyst where the support can be titania but preferably, is alumina for economic reasons. U.S. Patent 4,413,064 discloses an-alumina supported catalyst having cobalt, ruthenium and a Group IIIA or Group IVB metal oxide, e.g., thoria. European Patent 142,887 discloses a silica 8 supported cobalt catalyst together with zirconium, 8o88 titanium, ruthenium and/or chromium.

8 o e 8 SUMMARY OF THS INVENTION The invention resides in the preparation of a novel catalyst and the use of that catalyst in hydrocarbon synthesis reactions. The catalyst is comprised of cobalt and ruthenium, in intimate Z0 association, deposited on a titania support. Evidence suggests that atoms of cobalt and ruthenium are present in the same crystallite and that this intimate association of the metals provides the advantages mentioned hereinbelow.

The catalyst, when prepared as described herein, is an excellent hydrocarbon synthesis catalyst and may be used in hydrocarbon synthesis react'ions as other known catalysts are used, for example, as pellets loaded in tubes through which synthesis gas is passed and converted into higher hydrocarbons. The advantages of employing the particular cobaltruthenium catalyst of this invention in hydrocarbon synthesis are: lower methane yields and increased C 5 4 yields relative to a cobalt catalyst or a cobaltruthenium catalyst that has not been oxidized and re-reduced in accordance with this disclosure, greater cobalt time yields (that is, greater conversion of CO and H 2 per gram atom of cobalt per unit of time a measure of catalyst activity) and the ability to regenerate the catalyst, in situ, under low temperature flowing hydrogen. The last advantage differentiates from carbon burning operations which must take place at relatively high temperatures, e.g., I 400 0 C or higher in oxygen and, generally, requires **to *to removal of the catalyst from the reactor, an t.o expensive, time-consuming operation in commercial reactors.

.4 Ruthenium may promote hydrogenolysis and the intimate association of ruthenium with cobalt might t allow carbon deposits on the catalyst to be gasified via hydrogenolysis as opposed to carbon gasification via combustion with oxygen in cobalt catalysts other *i than those having the structure disclosed herein.

**e DESCRIPTION OF THE DRAWINGS Figure 1 shows the effect of intimate 4 4 association of cobalt and ruthenium on reduction temperatures as opposed to cobalt alone. The TG curve monitor's weight changes as the supported cobalt oxide is reduced in hydrogen from room temperature to 500 0

C

at 6 deg/min. The DTG plots the rate of weight change with time as a function of temperature. Figure 1 shows that the onset of reduction begins at a lower temperature with a calcined cobalt-ruthenium catalyst.

A cobalt-ruthenium catalyst not prepared in ^f 5 accordance with the procedures of this invention and wherein the cobalt and ruthenium are not intimately associated, reacts similarly as a cobalt only catalyst.

Figure 2 shows the effect of cobalt and ruthenium being in intimate contact on catalyst carburization, the tendency of carbon to grow on active sites of the catalyst as opposed to a cobalt only catalyst. Figure 2 follows the behavior of the catalysts heated from room temperature to 500 0 C in 4 ao 1:1 H 2 /CO following a prereduction. The large gain of weight between 3000 and 5000C results f om th growth of carbon. When the cobalt and ruthenium are in intimate contact, the growth of carbon is suppressed.

The cobalt only catalyst behaves similarly to a coba-lt-ruthenium catalyst wherein the cobalt and ruthenium are not in intimate contact, not precalcined.

Figure 3 and 4 show the results of traces of a cobalt-ruthenium catalyst prepared in accordance with this invention and developed from analysis with a high resolution transmission electron microscope with scanning transmission and energy dispersive x-ray analysis capabilities. Figures 3 and 4 show energy il dispersive x-ray traces (EDX) of calcined and uncalcined CoRU/Ti0 2 catalysts. The figures show that following the calcination and rereduction treatment the ruthenium has concentrated in the area 6 of the cobalt particle rather than remaining uniformly present throughout the support as it appears on the uncalcined, reduced catalyst.

DETAIELED DESCRIPTION In general, the hydrocarbon synthesis reaction is carried out at conditions that are known in the art. The H 2 :CO ratio is at least about 0.5 and up to about 10, preferably 0.5 to 4.0, and more E. opreferably about 1.0 to 2.5. The gas hourly space velocity can range from about 100 v/hr/v to about 5000 v/hr/v, preferably from about 300 v/hr/v to about 1500 v/hr/v and reaction temperatures may range from about 160 0 c to about 300 0 C, preferably about 190 0 C to 260 0

C,

while pressures are above about 80 psig, preferably about 80 to 600 psig, more preferably about 140 to 400 psig. Hydrocarbon synthesis results in the formation of hydrocarbons of carbon number range C 5 to aboit

C

4 0 or higher. Preferably, the synthesized hydrocarbons are primarily or almost completely paraffins.

The catalyst, cobalt and ruthenium on titania, contains about 5 to 25 wt.% cobalt, preferably 10 to 15 wt.% cobalt and about 0.03 to 0.30% ruthenium, preferably about 0.1 to 0.2 wt.% ruthenium. The atomic ratio of cobalt to ruthenium is about 10 to 400, preferably about 100 to 200.

Tho catalytic metals are supported on titania which may be used alone or with other inorganic refractory materials. Preferably, the support material is titania and more preferably the -7titania has a rutile:anatase ratio of at least about 2:3 as determined by x-ray diffraction (ASTM D3720-78) preferably about 2:3 to about 100:1 or higher, more preferably about 4:1 to 100:1 or higher, 100% rutile. The surface area cr: the support is, generally, less than about 50 m 2 /gm (BET).

Preparation of the catalyst is not believed to be a critical step insofar as deposition of the catalytic metals on the support is concerned. The intimate contact between the cobalt and the ruthenium is accomplished by subjecting the composition to an oxygen treatment subsequent to reduction of both of the metals. Consequently, the metals can be deposited impregnated) on the support either in serial fashion with the cobalt being deposited either before or after depositing the ruthenium or by co-impregnating the metals onto the carrier. In the case of serial impregnation, the carrier is preferably dried and the metal reduced prior to impregnation of the second metal after which drying and reduction is effected asain and prior to the treatment of the catalyst with an oxygen containing gas.

Preferabiy, the catalyst is prepared by depositing the cobalt, drying the catalyst, reducing the cobalt, depositing the ruthenium, also followed by drying and reduction, and to effect the intimate contact of the cobalt and ruthenium exposure to an oxygen containing gas, and a firial reduction.

8 Thus, the catalyst can be prepared by incipient wetness impregnation of the titania support with an aqueous solution of a cobalt salt, e.g., nitrate, acetate, acetyl acetonate or the like, the nitrate being preferred. The impregnated support is then dried and reduced in a reducing gas, such as hydrogen. Ruthenium is added to the reduced cobalt on titania catalyst using a ruthenium salt, e.g., ruthenium nitrate, chloride, acetylacetonate, carbonyl, etc. The catalyst is again dried and again reduced in a reducing gas, such as hydrogen. Intimate association of the cobalt and ruthenium is t accomplished by treating the reduced cobalt-ruthenium *11 I I t, on titania catalyst with an oxidizing gas, air I I or a dilute oxygen stream such as 20% oxygen in helium at elevated temperatures sufficient to oxidize the cobalt and ruthenium, for example above about 250 0 C, preferably 250 tJ 30 0 C; but not in excess of t about 600oC because of excessive oxide sintering. Upon ,t 1 reduction, the cobalt and ruthenium are intimately associated, that is, atoms of each are much closer '4 together than would otherwise be the case and are believed to be present in the same crystallite. Cobalt and ruthenium oxides in the bulk form a cabalt- S" ruthenium single phase mixed metal oxide, The available evidence suggests a likely bimetallic cluster fotmation of Co and Ru on the titania support.

Reduction is effected in hydrogen at about 400 0 C but can take place at temperatures ranging from about 200 to 500OC. Reduction of the catalyst is generally easier, thlcat is, occurs at lower temperatures, relative to a catalyst contailting only cobalt without tuthenium.

9 In virtually any catalytic process, catalyst activity decreases as run length increases due to a variety of factors: deposition of coke or carbon on the catalyst as a result of cracking, hydrogenolysis, or polymerization, buildup of poisons in the feed such as sulfur or nitrogen compounds, etc. In hydrocarbon synthesis reactions carbon tends to build up or grow (by complex polymerization mechanisms) on the surface of the catalyst, thereby shielding the catalytic metals from the reactants. Activity decreases and at some pre-set level of activity (as defined by conversion or selectivity or both), the process becomes sufficiently uneconomical to continue and the catalyst is either replaced or regenerated. In either case, downtime results and in the former, significantly increased catalyst costs are incurred.

Catalyst regeneration is desirable, particularly where regeneration can be accomplished withop't removing the catalyst from the reactor. Using the catalyst of this invention, regeneration can be effected by discontinuing the flow of carbon monoxide (and contir uing the flow of hydrogen if the gases are suLpplied separately) to the reactor or discontinuing the flow of synthesis gas (where synthesis gCj is the feed as produced, for example, by methane reforming or partial oxidation of methane) and flowing hydrogen to the reactor. After regeneration with hydrogen, synthesis gas flow to the reactor is resumed and the hydrocarbon synthesis reaction continued. The regeneration process may be conducted at intervals to return the catalyst to initial activity levels. Thus, a cyclical operation involving hydrocarbon synthesis and regeneration may be repeated.

10 The temperature in the reaction zone during hydrogen regeneration is preferably at or slightly above hydrocarbon synthesis reaction temperatures and pressures can be the same, as well; although neither temperature nor pressure are critical to the regeneration which is effected by the hydrogenolysis characteristics of the ruthenium bound intimately with the cobalt. In the case where the ruthenium is not intimately bound with the cobalt, not in the same crystallite, hydrogenolysis of the carbon deposited on the catalyst may have little or no effect on carbon deposited on the cobalt sites. Where the ruthenium and cobalt are in intimate association, ruthenium-promoted hydrogenolysis affects the carbon deposited on the particular crystallite and both t 'cobalt and ruthenium sites are regenerated, that is, freed of carbon deposits. It is only necessary that the conditions be conducive to hydrogenolysis promoted by ruthenium and carried out for a time sufficient to regenerate the catalyst. Preferably, temperatures range from about 150 0 C to about 300°C, more preferably about 190 0 C to 260 0 C and the hydrogen flow is continued until regeneration is effected, about 8 hours, preferably at least about 10 hours.

Regeneration results in the recovery of at least about 90%, preferably 95%, more preferably at least 100% of initial activity as measured by cobalt-time yields and is accompanied by C 5 yields greater than initially and CH 4 yields below initial yields. By "initial" We mean after the catalyst has stabilized, usually about 24 hours after startup, EXAMPLE 1: preparation and Evaluation of Supported Cobalt Catl-alysts Four cobalt-containing catalysts were prepared, three with titania as a support and one with silica. For catalyst A, 50 grams of Degussa titania was calcined at 5C\0 0 C for 4 hours X-ray dif fraction. showed that the titania contained rutile and 30% anatase; the LIET-measured surface area was 30M 2 /gm. 35 gins of coba~lt nitrate hexahydratpo CO(N0 3 2.6H 2 0 (Al~a, Pi atjonic Grade) were dissolved in 20 cc of doubly-distilled deiQnized. water. H-alf of the solution was impregnated by incipient wetness onto the titania. After the sample was dried at 100 0 C, the a remaining solution was impregnated onto the titania and the catalyst was dried at 1.00 0 for 1,6 hours.

Following calcination in air at 40QQC for 4 houro, the catalyst was placed in a tube furnace at 400 0 C in a hydrogen flow of 2000 cc HZ/ce Qat/hic foc a, period of 16 hours. After this reduction# He wi-,s introO.Qed for 2 hours and then a 1% stream of O~cygen~ Was added f.o the helium to paIssivate the co.t~lyst and allow its removaJl Iito the ambient environment* $Ubseqmont cobalt chemical analysis showed the cobalt aotl' be 11 Catalyst A the eote 00h~sts Ot 4 Co/TiQ 2 and Is desiqnate4 as QQ/T 102 1 n 0 following examiples.

For Catalyst B, 20 grams of 11.6% CO/TiQ 2 (4, portion of catalyst A) Were0 glec ted. 1.02 gramst of ruthenum nitrate, (hydtate) were d-Issolvod in of acetone. 20 grAmq~ of catalYSt A weret alutried into0 this solution and~ tho sOlVent was allowed to dvapdatiLe while beitig stigred. The catalyst was, dtiodi, 1rducedo and pasliVated 4s describc-d abov,,.

I

i i I.

I

SII(

Iir

I

12 To prepare Catalyst C, 10 grams of B were heated in 20% 02/80% He at 300 0 C for 4 hrs., rereduced in H 2 and passivated as described above. The cobalt and ruthenium Lontents in catalysts B and C were 11.6 and 0.14% respectively, corresponding to an atomic Co/Ru ratio of 160. The catalysts B and C are designated as CoRu/TiO 2 and CoRu(c)TiO 2 in the following examples.

Catalyst D, containing cobalt on silica, was prepared for comparison purpose.s. 30 grams of Davison 62 silica were calcined at 6000C for 4 hours. 50 grams of cobalt nitrate hexahydrate were dissolved in 40 cc of water. The solution was impregnated onto the silica. in four steps with intermediate dryings at 100 0 C, The catalyst was then dried, reduced and passivated as described above.

Chemical analyses indicated that the Co content was 23%. This catalyst is designated as Co/SiO 2 in the following examples.

EXAMPLE 2: Effect of Ru Promoter and Calcination at Low Pressures 5-10 cm 3 of catalysts A, 5, C, a id D from Example 1 were run in a single pass fixed bed reactor of 3/8 inch outer diameter. Hydrogen, carbon monoxide and nitrogen were obtained as a preblended mixture with 61+2% H 2 31+2% CO and 7+1% N 2 The feed mixture was passed over a Pd/Al20 3 catalyst (Deoxop Johnson Mathey), an activated charcoal sieve, and a 13X molecular sieve trap, to remove water, oxygen, and Ni and Fe carbonyls. Gas flows were controlled by Brooks mass flow controllers. Pressure 4 i'- 13was maintained with backpressure regulators.

Temperature was held isothermal to within 2 degrees by use of a Thermac temperature controller. Products were analyzed by capillary and packed column gas chromatography, using N 2 as an internal standard, C20-C200 molecular weight distributions were obtained by gas chromatography and gel permeation chromatography. Pretreated and passivated catalysts were rereduced in flowing hydrogen (200-4QO GHSV) at 400 0 C for 4 hours in the hydrocarbon synthesis reactor before Fischer-Tropsch experiments.

S.o Table I compares the Fischer-Tropsch WEo synthesis behavior of Co/TiO 2 (Catalyst A) with the E4 bimetallic CoRu/TiO 2 both directly reduced (Catalyst B) and calcined/rereduced (Catalyst C) as well as the comparative Co/SiO 2 catalyst (Catalyst D).

I, Hydrocarbon synthesis rates are reported as cobalt-normalized rates, cobalt time-yields, t defined as the moles of CO converted per hour per g-atom Co in the catalyst or as site-normalized rates (site-time yields) defined as the molecules of CO converted per hour per surface cobalt atom in the catalyst. The number of surface cobalt atoms is tI tdetermined from H 2 chemisorption measurements.

SHydrocarbon selectivities are reported on a carbon atom basis as the percentage of the converted CO which appears as a given product.

At 560 kPa the addition of Ru to Co/Tio 2 (CO/RU gm atom ratio 160) increases time yields more than threefold while decreasing CH 4 selectivity from 10.1% to Calcination of the bimetallic catalyst

^L

14 has a minor effect on selectivity, but it increases time yields by an additional 50%. Co/SiO 2 shows similar selectivities with about 50% higher time yield than Co/Ti0 2 because of the proportionately higher cobalt loading.

EXAMPLE 3: Catalysts A, B, and C were also compared at higher pressure, 2050 kPa, in the same reactor. Table

I*

II lists the results. At these conditions calcination *too. of the bimetallic Co-Ru/TiO 2 significantly improves performance. Time yields double with the addition of i+ r Ru to the Co/TiO 2 but improve an additional Sfollowing calcination. In addition, CH 4 selectivity S, decreases from 7.5 to 5.0% and the C 5 fraction increases from 86 to 91% following calcination and reduction.

EXAMPLE 4 Catalysts A, B, and C from Example 1 were run in a fixed bed reactor as described in Example 2 at 200 0 C and 560 kPa. During the run the conversions were varied between 5 and 70% by adjusting the space velocity between 200 and 3000 v/v/hr. Table III shows the CH 4 and C 5 selectivities as a function of CO conversion. For all three catalysts, the CH4 selectivity decreases and the C 5 selectivity increases with increasing conversion. At all conversion levels the methane yields are lower and C5+ yields higher for the Ru promoted catalysts. At all levels of conversion the calcination of the CoRu/TiO 2 catalyst decreases CH 4 and increases C5+ selectivities.

U -L i "~ft ft ft-ft ft..

ft ft ft.. -ft aft ft -ft 4 ft ft ft ft ft ft **ft ft Table 1.

Fischer-Tropsch Activities and Selectivities at 560 kPa

CO

Conversion CILA C1 5 Catalyst CO/SiO 2

(D)

Co/TiO 2

(A)

CoRu/TiO 2

(B)

CoRu/TiO 2

(C)

(Calcined) GHSV W t) Wt) Cobalt-Time Yield (h- 1 Space-Time Yield (h 1 450 300 1200 1800 28.9 27.7 26.0 25.3 8.4 10,.1 7.9 7.5 78.8 79.4 130 83 310 455 0.6 86.7 2.9 [200 0 C, H 2 /CO 2.05, 560 kPal hh.- A, A, a a 4 a Table 11.

Fischer-Tropsch Activities and !Selectivities at 20501

CO

Conversion CH 4 C5 Wt) Wt) 48.7 7.0 85-0 .cPa Catalyst Co/TiO 2

(A)

CoRu/TiO 2

(B)

CoRu/TiO 2

(C)

(Calcined)

GHSV

450 800 1200 Cobalt-Time Space-Time Yield Yield (h- 1 220 1.4 50.7 61.0 7.5 5.0 86.1 91.4 405 '730 2.6 4.7 [200 0 C, H 2 /CO =2.05, 2050 kPaj 41-4 TABLE III Fischer-Tropsch Activities and Selectivities as a Function of Conversion Catalyst Cobalt Time Yield CO conversion Cfl 4 Selectivity

C

5 Selectivity Co/Si0 2

(D)

1 1 7 65 9.5 7.4 75 82 CO/TiO 2

(A)

0.6 0.6 4 50 12 9.3 77.3 80.2 CORU/TiO 2

(B)

2.0 2.0 5 64 8.6 6.8 84.9 87.3 CoRu(c)/Ti0 2

(C)

2.9 2.8 5 68 7.8 85.3 87.8 Conditions: 200 0 C, 56OkPa, 11 2 /CO=2/l, conversion varied by changing space velocity 18 EXAMPLE Catalyst's A, B and C from Example 1 were run for periods of 10-30 days. During those time periods catalyst activity declines. Table IV shows the effect of hydrogen treatments on reactivating these catalysts.

Table IV.

Regeneration of Co Catalysts by H 2 Treatments *to* •r Cobalt- Time Yield CH4 (h- 1 (Wt%) Co/TiO 2 (4) Initial 0.6 8.9 80.1 Before H 2 treatment 0.5 9.5 81 After H 2 treatment 0.5 9.5 80.5 CoRu/TiO 2 (3) Initial 2.6 7.0 86 Before H 2 treatment 2.0 8.2 84 After H 2 treatment 2.6 6.5 87 CoRu/TiO 2 (calcined) Ir t Initial 4.5 5.5 91.0 Before H 2 treatment 3.9 6.4 88.8 i After H 2 treatment 4.8 4.9 91.5 24-48 hr. after H 2 H2 treatment at 200-230 0 C for 16 hr, 100 kPa Conditions, 50-60% CO conversion, 2060 kPa, 200 0 C, H 2 /CO 2/1 Conditions, 20% CO conversion 560 kPa, 200 2 C, H 2 /CO 2/1 For Co/Ti0 2 the CO conversion and CH 4 and C 5 selectivities do not respond appreciably to H 2 treatments, whereas the RU containing catalysts respond to the hydrogen treatment by regaining their -19 original activity and selectivity. For the calcined catalyst all resiults are superior to' izh results for the uncalcined ?,atalyst (B) EXAMPLE 6 The cal(7ined Co-Ru/TiO 2 catalyst (catalyst C) was run at two temperatures at a constant pressure of 2060 kPa. Space velocities were adjusted to keep conversion levels comparable. Table V presents the fli results Table V.

Effect of Temperature on Performance Of~ CORu(c)/TiO 2

(C)

Temperatures T/OC 184.8 200.0 G[HSV 600 1200 Co Conversion ()57.2 59.1 cobalt-Time yield (hV-1) 2.2 4.6 ECO/Kcal moll 21 ft carbon Selectivity

CH

4 3.4 5.4 ECH4(Kcal mol-1) 34

C

2 0 .40 0 .43

C

3 1.59 1.68

C

4 1.66 1.77

C

5 92.9 90.7 [CoRu(C)/TiO 2 2060 1kPa, H 2 /CO 2.05] 0.14% Ru, 11.6% Co 20 At higher temperatures selectivity to lighter products increases. The calcihed Ru promoted catalyst run at 0 C lower temperature has cobalt time yields comparable to the unpromoted Co/Ti02 and much higher

C

5 selectivity. Therefore, improved selectivities (less CH 4 and more C 5 are obtained at comparable metal yields.

EXAMPLE 7 Catalysts A and C were compared at different temperatures. Table VI lists the results.

*0 6 9 a 0 Table VI.

CoRu(C)/Ti0 2 Co/Ti0 2

(A)

Temp. 185 200 Co Time Yield (h- 1 2.2 1.4 93

CH

4 3.4 7 The data show that at similar cobalt time yields, the CoRu(c)/TiO 2 catalyst produces substantially more C 5 and less CH 4 than the Co/TiO 2 catalyst, the calcined catalyst being more active and mote selective to valuable products.

21 EXAMPLE 8 The Co/TiO 2 and CoRu(c)/Ti02 catalysts from Example 1 were treated under hydrogen in a thermalgravimetric analyzer (TGA) The samples were heated from room temperature to 500 0 C at 6 deg/min.

The TG curve monitors weight changes as the cobalt oxide is reduced to cobalt metal. The DTG plots the rate of weight change with time as a function of temperature. Figure 1 shows the onset of -eduction begins at a lower temperature with the CoRu(c)/TiO 2 catalyst. This indicates that the cobalt and ruthenium have come into intimate association on the t I t catalyst, Figure 2 shows the behavior of the Co/TiO2 and CoRU(c)/Ti02 c(atalysts in a 1:1 H 2 /CO mixture following reductior. The calcined CoRu/Ti02 catalyst does not grow carbon at temperatures where the noncalcined CoRu/TiO 2 or Co/TiO 2 do. Therefore, a combination of increased cobalt oxide reducibility and inhibited catalyst poisoning by carbon are believed to account for the increased number of active sites observed on calcined CORu/TiO 2 catalysts.

S

t EXAMPE 9 4 CoRu/Ti0 2 (catalyst B) and CoRu(C)/TiO 2 (catalyst C) were run under Fischer-Tropsch conditions for 700 hours, including two hydrogen regeneration treatments.

Electron microscopy studies of these catalysts were conducted using a Phillips EM-420ST high-resolution transmission electron microscope with scanning transmission and energy dispersive x-ray

TA

22 analysis capabilities. Under the conditions used in this study, the instrument had a resolution of better than 0.25 nm. The catalyst samples were ground using a mullite mortar and pestle and was ultrasonically dispersed in butyl alcohol. A drop of the suspension was then air dried on a carbon film.

Identification of the elements in the catalyst was made using the adjunct energy dispersive x-ray (EDX) analyzer. Using the EDX system, particles as small as I nm were analyzed. With these catalysts and with a 1 nm beam for analysis, the x-ray spatial resolution was approximately 2.5 nm.

Detectability limits for the elements in question were about 0.3 0.4 weight percent in the volume analyzed.

t Figure 3 shows the results.

The morphology of the cobalt particles on titania is similar on both monometallic and bimetallic catalysts. Cobalt is dispersed on the titania as slightly elliptical particles 20-50 nm in size. EDX analysis of these particles suggests that ruthenium is present with the cobalt in the same crystallite after calcination and reduction treatments. Figure 3 shows that following the calcination and rereduction treatment the ruthenium has concentrated in the area of the cobalt particles so that ruthenium above detectability limits was not observed on the titania, but was only in the cobalt particles. (In the un,-alcined CoRu/T10 2 (Figure ruthenium was below detection limits on the support and in the cobalt particles, indicating that Ru wa~s not preferentially concentrated, but remained uniformly present.) 1-0k*0*

Claims (9)

1. A hydrocarbon synthesis catalyst comprising catalytically active amounts of cobalt and ruthenium on a refractory support comprising titania having a surface area of less than 50 m 2 /g and characterised in that atoms of cobalt and ruthenium are disposed within the same crystallite, by reducing both cobalt and ruthenium prior to exposure to an oxygen containing gas, forming the metal oxides, and reducing the oxides. V.

2. The catalyst of claim 1 wherein cobalt is ou present in amounts ranging from 5 to 25 wt.% of the catalyst and the atomic ratio of cobalt to ruthenium is 10 to 400.

3, A process for preparing a hydrocarbon synthesis catalyst which comprises impregnating a refractory support comprising titania having a surface area of less than 50 m /g with catalytically active amounts of cobalt and ruthenium salts, drying the impregnated support, reducing the cobalt and ruthenium, treating the reduced metal with an oxygen containing stream at conditions sufficient to form oxides of cobalt and oxides of ruthenium, and reducing the cobalt and ruthenium oxides,

4. The process of claim 3 wherein the cobalt and ruthenium are co-impregnated onto the support, r ct f

5. The process of claim 3 wherein the cobalt is first impregnated onto the support, dried and reduced in hydrogen and then the ruthenium, 4. impregn.ted onto the support, dried and reduced in hydrogen. irfA.-- i 24

6. The process of 01,a m wherein the re- duced metals are treated with a, o~qygen containing stream at a temperature above 464,- 2 ,o0C for a period sufficient to form cobalt oxide and r ohenium oxide or a bimetallic cobalt-ruthenium oxide,

7. The process of claim 6 wherein the cobalt and ruthenium oxides are reduced in the presence of hydrogen at temperatures ranging from -a-boat-200 0 °C to -a bou-500OC.

8. A hydrocarbon synthesis process which comprises reacting synthesis gas in the presence of a catalyst comprised of titania wherein atoms of cobalt and ruthenium are disposed within the same 'crystallite, at reaction conditions suitable for the formation of higher hydrocarbons.

9. The process of claim 8 wherein the hydrocarbon synthesis process is intermittently interrupted, synthesis gas feed to the catalyst is discontinued and the catalyst is regenerated in the presence of hydrogen. The process of claim 9 wherein regeneration is effected at temperatures ranging from 160 0 C to *at -b -4b-00 0 o an and at least of the catalyst's initial activity is recovered. DATED this 11th day of December 1987. EXXON RESEARCH AND ENGINEERING COMPA 4, EDWD., WATERS SONS VPATENT ATTORNEYS FM50 QUEE STREET Mt BOURNE. VIC. 3000.