Interaction of Matrix and Zeolite in Catalytic Cracking
Volume 5 Issue 1

Introduction

Modern fluid catalytic cracking (FCC) catalysts are composed of crystalline zeolite Y which is surrounded and held together by an amorphous silica/alumina matrix. Matrix contributes significantly to the overall performance of FCC catalysts. Many of the functions of the catalyst matrix such as upgrading bottoms, reducing SOx emissions, and passivating catalyst poisons have been described previously (Ref. 1-4). In this paper we will illustrate how matrix can interact with the zeolite in the cracking process to affect conversion and selectivity. It will be shown that the catalyst matrix can be tailored to provide refiners with desired product quality and quantity. Therefore, in selecting FCC catalysts refiners should understand the type of matrix they need for their particular application.

Synergy of Matrix and Zeolite Increases Conversion

To demonstrate how matrix and zeolite interact to crack gas oil molecules, catalysts with varying matrix types were prepared and their selectivities measured using BASF's micro-activity test (MAT) unit. All MAT runs used BASF's standard mid-continent gas oil, a catalyst-to-oil ratio of 5, a weight hourly space velocity of 15 and a reactor temperature of 910F (488C)(Ref. 5). In this article a catalyst's activity is reported either as a MAT conversion, which is the weight percent of gas oil converted to products boiling below 421F (216C), or as a dimensionless activity based on second order kinetics, where Activity = Conversion /(100-Conversion). The MAT products gasoline, light cycle oil (LCO), and bottoms are defined as the fractions to 421F (216C),421 to 602F (or 216 to 317C), and >602F (317C), respectively. Fresh catalysts were steam deactivated at temperatures of 1350F (732C) or 1450F (790C) for periods of 4 to 8 hours prior to testing. Within each set of experiments, the same steaming severity was used for each catalyst.

In the first set of experiments, the cracking activities of three catalysts were determined. Catalyst 1 was a rare earth exchanged Y (REY) zeolite catalyst with minimal matrix activity and Catalyst 2 was a moderate activity amorphous catalyst.

MAT conversions of Catalysts 1 and 2 were 52 wt.% and 30 wt.%, respectively. The zeolite catalyst's activity (1.1) was nearly three times higher than that of the amorphous catalyst (0.4) . A third catalyst was prepared from a blend of the first two so that it had the same zeolite activity as the first plus the same amorphous activity as the second. If the effects of matrix and zeolite were additive, the activity of Catalyst 3 would be calculated to be 1.1+0.4 = 1.5. In fact, when the activity was measured, it was found to be 50% higher than the calculated value. The experimentally determined MAT conversion of Catalyst 3 was 69 wt.%, which is equal to an activity of 2.2. These results are summarized in Figure 1.

The fact that the measured activity of Catalyst 3 is substantially greater than the sum of the activities of the other two catalysts indicates that there is significant interaction between the components. This synergistic interaction can be explained by a mechanism of initial matrix cracking of large feedstock molecules to smaller ones, and subsequent zeolite cracking of the smaller molecules to converted products.

The mechanism of primary matrix cracking of hydrocarbons followed by secondary zeolite cracking can be better understood when the zeolite pore size is considered. The zeolite pore size is not suitable for cracking large hydrocarbon molecules since the pores are too small (< 8 Angstroms in diameter) to allow them to diffuse to the cracking sites. The zeolite catalyst in this experiment, for example, can convert only about half of the feed to products having a boiling point below 421F (216C). With the addition of the amorphous catalyst which contains larger pores ( >50 Angstroms) however, many additional molecules can be cracked on the more accessible matrix cracking sites. Relatively few of the products of matrix cracking are in the gasoline boiling range or lighter, so by the above definition, the amorphous catalyst provides minimal "conversion". The amorphous catalyst does, however, produce a wide range of intermediate, partially cracked products whose boiling points exceed 421F (216C ). These intermediate products, when added to the lighter feed molecules that are already of a size allowing diffusion into the zeolite, significantly increase the fraction of hydrocarbon molecules which are available for zeolite cracking. The result is a synergistic interaction between matrix and zeolite in which the activity attained by their combined effects is far greater than the sum of their individual effects.

Staged MAT Reactor Experiment Determines Effects of Sequential Cracking

"Staged bed" MAT experiments can be used to illustrate the interaction of matrix and zeolite even more clearly. A second experiment was run using a 2 stage MAT reactor bed, where the stages were layers of different catalyst types. The same feedstock and catalysts were used in these staged-bed experiments as were used in the previous single-bed experiments.

In the first experimental run, the staged bed MAT reactor had a lower layer of the high zeolite, low matrix activity catalyst and an upper layer of the moderately active amorphous catalyst. With the oil vapors flowing downward through the reactor, the oil contacted the amorphous catalyst first and the zeolite catalyst second. In the second run, the catalyst layers were reversed so that the oil vapors first contacted the zeolite catalyst and then the amorphous catalyst. Both runs used the same reaction conditions and equal amounts of catalyst. The results of the staged bed MAT runs are shown in Table 1. They indicate that the "matrix-first" configuration has a large conversion advantage over the "zeolite first" run. With the "matrix-first" configuration, a conversion of 69 wt.% was achieved. With the catalyst beds reversed, the conversion dropped to 51 wt.%. Compared on an activity basis, the "matrix-first" catalyst arrangement had more than twice the activity (2.2) of the other arrangement (1.0). Other investigators have observed similar improvements in MAT conversion when an amorphous catalyst preceded a zeolite catalyst(Ref. 6).

The results of the staged bed MAT runs demonstrate more clearly the interaction of matrix cracking and zeolite cracking. With the matrix bed ahead of the zeolite, conversion improved by 18 wt%, with about 80% of the additional conversion occurring by the cracking of bottoms. The bottoms molecules that were unconverted in the "zeolite first" run are upgraded to valuable products (gasoline and LCO) in the "matrix first" run. The selectivities of converted products are also better when the matrix contacts the feed first and precracks it. In the "matrix first" run, dry gas selectivity declined 25%, and coke selectivity declined 36%, relative to the "zeolite first" run.

Comparing the zeolite catalyst results in Figure 1 and the "zeolite first" staged bed results in Table 1 shows that the MAT conversions achieved in the two experiments are nearly equal. The matrix bed of the staged bed experiment does not enhance cracking of gas oil which has already been cracked by the zeolite. This suggests that the matrix can promote primary cracking of relatively large hydrocarbon molecules such as those found in a fresh feed, but it is inactive to smaller, less reactive hydrocarbons such as those generated from zeolite cracking.

Matrix Type Affects Selectivities

Matrix acidity can be characterized using various techniques. The acid type, strength, and performance characteristics of two different matrix systems having comparable surface areas were distinguished by taking infrared (IR) spectra of sorbed pyridine on matrices(Ref. 7), by a n-hexadecane cracking test reaction(Ref. 8), and by MAT studies. The first method uses infrared spectroscopy to determine how much of a basic material such as pyridine remains on an acid catalyst at elevated temperature relative to the amount adsorbed at a lower temperature. This gives a measure of the strength of acid sites since a stronger acid will require a higher temperature to break the acid-base interaction, and will therefore retain more pyridine. Study results demonstrated that matrix S had much greater pyridine retention at elevated temperatures than matrix W, indicating that it contains much stronger acid sites than matrix W. As shown in Figure 2, matrix S retained approximately 70% of the adsorbed pyridine at 450C(842F) as compared to only 20% for matrix W. Model compound MAT test results using n-hexadecane (C16H34) as feed showed 70% higher olefins and aromatics formed for matrix S than for matrix W, indicating that a strongly acidic matrix increases gasoline octane. These results are summarized in Table 2. A MAT comparison of matrix S and matrix W using standard gas oil showed that matrix W had only half the cracking activity of matrix S but better liquid selectivities. These results are summarized in Table 3.

How matrix acidity and surface area can be manipulated to optimize cracking selectivity and octane is illustrated in the final set of experiments. In these studies three catalyst formulations, each with a different tape of matrix and having REY as the zeolite component, were tested in the MAT unit using the normal single stage reactor configuration. The matrix types that were used varied in composition, surface area and acidity of cracking sites. The MAT yields of the three catalysts at 70 wt.% conversion are shown in Table 4. The term "matrix activity" used in the table refers to the combined effects of matrix acidity and surface area.

The test results show that significantly different selectivities result depending on the type of matrix used. Catalyst A, which has the lowest matrix activity, has the highest yields of coke and gas and the lowest yields of the most valuable liquid products (gasoline and LCO). This is because the few matrix sites in Catalyst A are strongly acidic and promote overcracking of gasoline. Catalyst B, with its moderately active matrix, has the lowest coke and gas yields. It also provides much improved yields of valuable products compared to Catalyst A. Catalyst C, which has the highest matrix activity, produces yields that generally fall between those of Catalyst A and Catalyst B. These results suggest that there is an optimum level of matrix activity for each situation, below which or above which valuable products will be degraded to less valuable products.

MAT paraffin-to-olefin ratios of the C4's are included in Table 4 to indicate the relative octane trend of the three catalysts. Higher gasoline octane is expected when the iC4/C4= ratio is lower. On this basis Catalyst C would have the highest gasoline octane, followed by Catalyst B, and then Catalyst A. Because Catalyst A produced higher yields of coke and gas, lower liquid products, and poorer octane, we conclude that a low activity matrix offers no performance benefits. In contrast, the catalysts with higher matrix activity, Catalysts B and C, provide varying degrees of improvement in selectivities and product quality.

Conclusion

FCC catalyst matrix plays a key role in determining overall activity, selectivity, and octane trends. Matrix provides primary cracking sites, generating intermediate feed molecules for further cracking to desirable products by the internal zeolite sites. The type of matrix used in the cracking catalyst alters activity and selectivities and can be controlled by the manufacturer to meet specific refining needs.

References

1. Silverman, L.D.; Winkler, W.S.; Tiethof, J.A. and Witoshkin, A. Paper AM-86-62, N.P.R.A. Annual Meeting, March 23-25, 1986.

2. Otterstedt, J.E.; Zhu, Y.M. and Sterte, J. Appl. Catal., 38, 143-55 (1988).

3. Otterstedt, J.E.; Gevert, S.B.; Jaeraas, S.G. and Menon, P.G. Appl. Catal., 22, 159-79 (1986).

4. Scherzer, J. and McArthur, D.P. Ind. Eng. Chem. Res., 27, 1571-6 (1988).

5. Speronello, B.K. and Reagan, W.R. Oil Gas J., 82, 139 (1984).

6. Ritter, R.E. and Young, G.W. Paper AM-84-57, N.P.R.A. Annual Meeting, March 25-27, 1984.

7. Parry, E.P., J. Catal., 2, (1963), 371.

8. Gladrow, G.M.; Krebs, R.W. and Kimberlin, C.N. Ind. Eng. Chem. Prod. Res. & Dev. 45 (1953) 142-147.

[An abbreviated version entitled "FCC Matrix/Zeolite Interactions" was published in the February 1990 issue of HYDROCARBON PROCESSING, pp 55-56.]