NON-SULFIDED Ni-BASED HYDROCRACKING CATALYSTS

Field of the Invention

The invention relates to hydrocracking catalysts.

Background to the Invention

The Low-Temperature Fischer-Tropsch (LTFT) process includes within its primary products a significant volume of heavy hydrocarbons, collectively referred to as waxes. Conventionally, these waxes, which are essentially free of sulfur, are hydroconverted via hydrocracking reactions into distillates.

As indicated by J Scherzer and AJ Gruia in "Hydrocracking Science and Technology" (Marcel Dekker, 1996), commonly used commercial hydrocracking catalysts are those based on NiW, NiMo and CoMo on amorphous silica-alumina systems. All these catalysts demand the continuous addition of a sulfur-containing species in order to maintain their performance.

The use of amorphous silica-alumina as support may produce higher selectivities to distillates, this being a consequence of its lower acid strength, in contrast to that found in the strongly acidic zeolitic supports (Calemma et a/., Studies in Surface Science and Catalysis, 136 (2001 ) 302).

Through the use of such hydrocracking processes, the advantage of starting with a sulfur-free feed and finishing with a sulfur-free product will be lost, and at the same time, H ₂ S will be present in the tail gas. Most importantly, however, it will be highly advantageous in terms of the process economics to develop a non-sulfided non-noble metal catalyst.

Ciapetta and Hunter, in Industrial Engineering and Chemistry, 45 (1953) 147, reported on the use of a non-sulfided NiZSiO ₂ -AbO ₃ catalyst for the

hydrocracking of n-hexane and n-octane. It is, however, well known in the technical literature (see for example Lugstein et a/, in Applied Catalysis A: General, 152 (1997) 93), that supported Ni catalysts exhibit high hydrogenolysis activities, resulting in the production of methane, an undesirable low-value product in most cases. This observation applies to Ni supported on all the commonly used supports such as silica, alumina, silica- alumina, zeolites, and even basic supports, e.g. magnesium oxide. Moreover, formation of methane has to be minimized since it influences the hydrogen partial pressure in a hydrocracker operating in a gas recycle mode.

Due to commercial reasons, in the conversion of valuable carbon-containing species, it is also desirable to minimize the consumption of hydrogen in the production of the less valuable product, viz. methane.

From the above, it can be gathered that a need exists for a non-sulfided non- noble metal F-T wax hydrocracking catalyst of low hydrogenolysis activity.

Summary of the Invention

Surprisingly, the inventors have found a workable formulation for an active and selective non-sulfided Ni-based catalyst of low hydrogenolysis activity and long lifetime (avoidance of loss of the dehydrogenation/hydrogenation activity due to metal sintering etc), whereas the technical literature pertains predominantly to non-sulfided noble metal or sulfided NiW, CoMo or NiMo on zeolitic or non-zeolitic acidic supports.

According to one aspect of the invention, there is provided a hydrocracking catalyst, which catalyst is non-sulfided and has a Ni content of at least 1 mass% and a silica content of at least 20 mass%.

The catalyst may have a Ni content of at least 3 mass%.

The catalyst may have a Ni content of at least 4.5 mass%.

The catalyst may have a Ni content of up to 50 mass%.

In some embodiments, the Ni content is between 5 mass% and 12 mass%, typically in the range 6 mass% to 10 mass%.

The catalyst may have at least 40 mass% silica.

The catalyst may include in excess of 60 mass% silica, and even in the region of 80 mass% silica, or even up to 99 mass% silica.

The silica may be present in the form of silica-alumina.

The catalyst may include Sn.

The catalyst may include more Ni than Sn.

Typically, the catalyst may include Ni and Sn wherein the Ni:Sn molar ratio exceeds 1 :1.

The Ni:Sn molar ratio may exceed 2:1 , 3:1 , or even higher.

In one embodiment the Ni:Sn molar ratio is 6:1.

According to another aspect of the invention, there is provided a method for reducing methane formation when hydrocracking hydrocarbons in the presence of a non-sulfided Ni containing catalyst.

The catalyst may contain Sn.

The silica may be present in the form of silica-alumina.

The method may reduce the selectivity to methane to below 0.13 mass%,

typically to 0.011 mass% or less, and even to 0.008 mass% or less.

The method may include including Sn in a quantity such that the molar ratio of Ni:Sn is in excess of 1 : 1.

The method may include including Sn in a quantity such that the molar ratio of Ni:Sn is in excess of 5: 1.

The method may include using a catalyst which has in excess of 3 mass% Ni, typically in excess of 4.5 mass% Ni, preferably in excess of 5 mass% Ni.

The method may include using silica-alumina as the support for the catalyst.

According to a further aspect of the invention, there is provided a process for the hydrocracking of hydrocarbons, said process including exposing said hydrocarbons, for example paraffinic hydrocarbons boiling in the 37O ⁰ C+ range, also referred to as waxes, or primary F-T derived waxes, to a catalyst as described above in a reactor operating at hydrocracking temperatures and pressures.

The process may also be used for the hydroconversion of lower boiling hydrocarbons, such as naphtha or middle distillates derived from an F-T process.

The process may be performed in the temperature range of 200-450 ⁰ C, at a pressure of 5-250 bar, and a Weight Hourly Space Velocity (WHSV) range of 0.1-10 h- ¹ .

The nickel-tin (NiSn) catalyst may also be used in a process for the hydrocracking of crude oil fractions, bio-mass, and in general, any source of available hydrocarbonaceous material.

The formation of methane may be reduced by using the non-sulfided Ni-

based hydrocracking catalyst to less than 1 mass%, typically less than 0.1 mass%.

With the use of Sn in the catalyst, the methane formation may be reduced to less than 0.03 mass%, preferably to less than 0.01 mass%.

The silica may be present in the form of silica-alumina.

The methane formation, or methane yield, is calculated as the selectivity multiplied by the fractional conversion.

Specific Description and Examples

A series of Ni-based catalysts was synthesized using a commercial silicated alumina having a silica content of 40 mass%.

The 3 and 4.5% Ni/silicated alumina samples were prepared by wet impregnation using aqueous solutions of nickel nitrate of the appropriate concentration to give the desired metal loading. These samples were tested as catalysts for the hydrocracking of n-hexadecane, which was used as the model compound. The product distributions obtained over these catalysts, which were non-sulfided, appeared symmetrical, and were close to ideal hydrocracking, as defined by J Weitkamp and S Ernst in "Guidelines for Mastering the Properties of Molecular Sieves", Plenum Press, 1990, p. 343. In the hydrocracking of heavier hydrocarbons, such as F-T waxes, ideal hydrocracking implies that the desired distillates selectivity will be at its theoretical maximum. The results obtained, therefore, are positive indicators that non-sulfided Ni is a suitable metal in serving as the dehydrogenation/hydrogenation function in hydrocracking catalysts. However, it was also observed that the amount of methane produced was significant.

A number of patents dealing with supported sulfided nickel-tin hydrocracking

catalysts were granted to the Chevron Research Company in the period of 1968-1970 (US Patents 3,399,132 (1968), 3,542,696 (1970), and 3,598,724 (1971 )). The objective of tin addition to the nickel catalysts was, as indicated in these patents, to increase the activity of the catalysts. However, the influence of tin on hydrogenolysis was never noted since sulfiding completely eliminates hydrogenolysis.

The inventors prepared a series of Sn-containing Ni-based catalysts for use in hydrocracking. A NiSn/silicated alumina sample was prepared via the co- impregnation of a Sn compound using the molar ratio of Ni:Sn = 3:1. This supported NiSn catalyst was then tested for the hydrocracking of n- tetradecane.

The catalyst without Sn (i.e. the 7%Ni/silicated alumina) produced both a highly symmetrical hydrocracking product distribution as well as a significant quantity of methane, about 20 mole%. It was surprisingly found, however, that the addition of the Sn to the silicated alumina-supported Ni-based hydrocracking catalyst resulted in the almost total elimination of the hydrogenolysis activity of the catalyst (these results are shown in Figure 2 in Example 2). Such an observation has not been made previously in the patent or open literature on non-sulfided Ni-based hydrocracking catalysts.

From the above experiments it appears, therefore, that we have been able to overcome the problem of hydrogenolysis by the addition of tin to the nickel- based hydrocracking catalysts. A problem found with the Ni or NiSn/silicated alumina catalysts, however, was the stability of the catalyst with time-on- stream (TOS), namely, the loss of metal function and the shift to lighter products. While initially in the run almost ideal hydrocracking was obtained, after several days on-stream, a shift to lighter products was observed. These observations are demonstrated in Example 3.

From comparative Temperature-Programmed Reduction studies of nickel oxide supported on reference supports such as silica, alumina and silica-

alumina, it was concluded that in the silicated alumina-supported nickel catalyst precursors, the nickel oxide is preferentially associated with the alumina phase. Since the loss of metal function could be a consequence of this association, we focused our attention on commercial silica-alumina samples which had to have a low content of alumina (typically used as binder), a high silica content but also a high tetrahedral aluminium content in the silica-alumina phase in order to possess high Brønsted acidity. As these catalyst properties are not obtainable from the manufacturers, a catalytic test reaction (dehydration of 1-hexanol) was used to ascertain the acidic activity of the silica-alumina extrudates. To achieve this, the reaction temperature was kept low (200 ⁰ C) in order to minimize the contribution of the alumina phase present in the silica-alumina extrudates to the dehydration activity, and thereby observe predominantly the catalytic activity of the silica-alumina (and of its tetrahedral aluminium content). To better understand the above, reference should be made to Table 1 in Example 4 which gives the dehydration results obtained with the different commercial products.

Following the above studies, another series of Sn-containing Ni-based catalysts for hydrocracking was also prepared using the commercial silica- alumina which contained 50 mass% silica (designated as SA2 in Table 1 ) and exhibited a high acid catalytic activity (e.g. dehydration of the alcohol to the hexenes). The Ni content was varied from 6 to 10 mass% and a 6:1 mole ratio of Ni:Sn was used. These catalysts were tested in bench-scale reactors for periods of up to 600 hours, using again n-tetradecane as the model compound. From the results obtained it could readily be ascertained that the addition of Sn almost completely eliminates the degree of hydrogenolysis (0.008 mass% selectivity to methane). The beneficial effect of Sn addition for the suppression of methane formation in hydrocracking reactions is clearly evident from these examples as well (see results in Tables 3 and 4 in Example 5).

However, due to the inclusion of Sn, the molar distribution of the cracked products had shifted to lighter products and more Ni would have to be added

to obtain a better balance between the acid and metal functions and hence attain ideal hydrocracking. Also, optimization of a hydrocracking catalyst in terms of the % Ni and the Ni:Sn ratio has to be determined for each catalyst system.

Furthermore, whereas with the catalyst prepared using the silicated alumina as the carrier, a shift was already clearly noticeable after 300 hours on- stream, the NiSn/silica-alumina catalysts were stable under the same operating conditions and produced similar conversions and product distributions with TOS.

It was also found with these catalysts that with increasing metal loading there was also a slight shift to higher carbon numbers in the product distribution. These results clearly show that higher metal loadings are required for this particular silica-alumina in order to obtain the appropriate balance between the metal and acid functions.

It was also again surprisingly found that the methane formation over the Ni/silica-alumina was considerably less than that obtained over the Ni/silicated alumina. It can be concluded, therefore, that the higher the silica content of the support/acidic component, the lower the selectivity to methane. These comparative and unexpected results are listed in Table 5 of Example 5.

Example 1

The 3 and 4.5%Ni/silicated alumina were prepared using aqueous solutions of nickel nitrate hexahydrate (99% pure, Aldrich) of the appropriate concentration in order to achieve the indicated % metal loading (taking into account that the Loss on Ignition = 13.8 mass%). The solvent was removed using a rotary evaporator at 50 mbar and 55 ⁰ C. This was followed by drying at 120 ⁰ C overnight and calcination at 300 ⁰ C for 2 hours. After loading a sample in the reactor, in-situ reduction was carried out at 400 ⁰ C for 16 hours using hydrogen at atmospheric pressure. The samples were then tested as

catalysts for the hydrocracking of n-hexadecane (n-Ci ₆ ), which was used as the model compound. The reaction conditions for the 3%Ni/silicated alumina were 350 ⁰ C, 55 bar, WHSV = 2.3 h ^"1 and an H ₂ //7-C ₁₆ mol ratio of -10, and for the 4.5%Ni/silicated alumina were 345 ⁰ C, 55 bar, WHSV = 2.5 h ^"1 and an H2/A7-C16 mol ratio ~9. The product distributions obtained at -41% conversion over these catalysts, which were non-sulfided, appear symmetrical, and are therefore close to ideal hydrocracking.

Example 2 A NiSn/silicated alumina sample was prepared via the co-impregnation of a tin compound using the molar ratio of Ni:Sn = 3:1. This sample was prepared by dissolving 13.5 g of Ni(NO ₃ ) ₂ -6H ₂ O and 3.5 g of SnCI ₂ -2H ₂ O (Aldrich) in 150 ml of 95% ethanol. To this solution, 40 g of the silicated alumina support was added and the mixture was allowed to stand for 1 h at room temperature. The solvent removal and drying steps were carried out as described in Example 1 followed by calcination at 600 ⁰ C for 3 h. After reduction at 450 ⁰ C for 16 h, this supported NiSn catalyst was then tested for the hydrocracking of /7-tetradecane (π-C14) at 31 mass% conversion under the reaction conditions of 340 ⁰ C, 50 bar, WHSV = 1.5 h ^"1 and an H ₂ /π-C ₁₄ mol ratio of -10. The product distributions obtained over the Ni and NiSn/silicated alumina are shown in Figure 2 which demonstrate clearly the beneficial effect of tin addition to the nickel-based hydrocracking catalyst for the suppression of hydrogenolysis.

Example 3

A silicated alumina-supported NiSn catalyst was also prepared containing 5% Ni and 1.7% Sn by mass using the same procedure described in Example 2. The catalyst precursor was calcined at 350 ⁰ C for 2 h, reduced at 350 ⁰ C for 4 h, and the catalytic reactions were carried out at 343 ⁰ C, 50 bar, WHSV=1.8 h ^{" 1} and using an H ₂ //7-C-ι ₄ mol ratio of -10. The product distributions obtained at 16 and 514 hours on-stream are shown in Figure 3. The shift to lighter products with TOS due to the loss of metal function is clearly evident.

Example 4

Table 1. The use of 1-hexanol dehydration for the evaluation of the Brønsted acidity of commercial silica-aluminas Reaction conditions: 200 ⁰ C, WHSV = 3.0 h ^"1 and TOS = 1.0 h

Neat silica-alumina prepared using a literature method (J Heveling, CP Nicolaides and MS Scurrell, Applied Catalysis A: General, 173 (1998) 1 ).

Commercial silica-alumina samples.

Example 5(a)

A second series of NiSn catalysts was prepared using the high-acidity SA2 silica-alumina as support, which had a silica content of 50 mass%. The Ni content was varied from 6 to 10 mass% and a 6:1 mol ratio of Ni:Sn was used. All samples were calcined at 350 ⁰ C, and 10 ml of catalyst precursor diluted with 10 ml of carborundum were loaded in the reactor. Reduction was performed for 16 h at 350 ⁰ C under atmospheric pressure using a hydrogen flow of 20 IN/γI. Table 2 shows the basic characteristics of the catalysts.

Table 2. Characteristics of the NiSn/silica-alumina catalysts

Example 5(b)

The catalysts described in Example 5(a) were used for the hydrocracking of n-tetradecane under the reaction conditions listed in Table 3. The % conversions and % methane selectivies obtained are also given in the same Table.

The methane selectivity had decreased from about 0.13 mass% with the 7%Ni/silica-alumina sample to 0.008 mass% with the Sn-containing catalyst (see Tables 3 and 4). The beneficial effect of Sn addition for the suppression of methane formation in hydrocracking reactions is clearly evident from these examples as well.

Table 3. Hydrocracking of /7-tetradecane over the catalysts of different % Ni and Sn loadings ¹

Reactions carried out in bench-scale reactors using 10 ml of catalyst diluted with 10 ml of carborundum.

Example 5(c)

The % conversion and % selectivity to methane as a function of TOS for the 10%Ni, 3.4%Sn/silica-alumina catalyst are listed in Table 4. The reaction conditions were 323 ⁰ C, 50 bar, and an H ₂ Zn-Ci ₄ mol ratio of -10. The results show the stable performance of the catalyst with TOS as well the extremely

low levels of methane formation.

Table 4.

% Conversion and % methane selectivity as a function of time-on-stream for the 10% Ni, 3.4% Sn/silica-alumina catalyst

Example 5(cQ The product distributions obtained at different times on-stream over the catalyst described in Example 5(c) are shown in Figure 4. The reaction conditions are as given in Table 3. The results show that similar product distributions are obtained with this catalyst at different times on-stream and that there is clearly no shift to lighter products.

Example 5(e)

Figure 5 shows the product distributions obtained over the catalysts with different metal loadings, as described in Table 3. It can be seen that there is an increasing shift to higher carbon numbers (C _θ -CH) with increasing metal loading.

Example 6

In this example, a comparison is made between the % selectivities to methane observed in the experiments conducted using the catalysts

containing nickel only and the two different supports. The results clearly show that even though a higher nickel loading was used in the Ni/silica-alumina catalyst, a much lower selectivity to methane was attained. The beneficial effect of a high silica content and a low alumina content in hydrocracking catalysts for minimizing the extent of methane formation (hydrogenolysis) is thus demonstrated.

Table 5. Effect of alumina content on methane selectivity