Unit Cell Volume as a Measure of Dealumination ... - ACS Publications

Sep 23, 2014 - Jürgen Knöll†, Udayshankar Singh‡, Jeffrey Nicolich§, Rafael Gonzalez†, Mike Ziebarth‡, Christoph Fougret†, and Stefan Brandt†...
1 downloads 0 Views 539KB Size
Article pubs.acs.org/IECR

Unit Cell Volume as a Measure of Dealumination of ZSM‑5 in Fluid Catalytic Cracking Catalyst Jürgen Knöll,*,† Udayshankar Singh,‡ Jeffrey Nicolich,§ Rafael Gonzalez,† Mike Ziebarth,‡ Christoph Fougret,† and Stefan Brandt† †

Grace GmbH & Co. KG, In der Hollerhecke 1, 67547 Worms, Germany W.R. Grace & Co.-Conn, 7500 Grace Drive, Columbia, Maryland 21044, United States § W.R. Grace & Co.-Conn, 62 Whittemore Avenue, Cambridge, Massachusetts 02140, United States ‡

ABSTRACT: A fluid catalytic cracking (FCC) catalyst is used in refineries to upgrade crude oil into valuable products. To fulfill the increasing demand for propylene, additives are used to boost propylene production. The hydrothermal conditions in a FCC unit gradually deactivate the Y-zeolite in the catalyst and the ZSM-5 zeolite in the additive by a dealumination of the framework structure. By powder X-ray diffraction, the change in the lattice parameter of the Y-zeolite over time is used to monitor the deactivation process. In this paper, we have shown that with a combination of powder X-ray diffraction and Rietveld analysis we can monitor the deactivation of ZSM-5. The lattice parameters measured for ZSM-5 are used to calculate the unit cell volume, which shows a good correlation with physical properties, such as total acidity and surface area, and the catalytic properties of ZSM-5 based additives. The data demonstrate that the technology can be applied to both laboratory and field deactivated ZSM-5 based additives.

1. INTRODUCTION Fluid catalytic cracking (FCC) is the most important process in the refining industry to upgrade high boiling hydrocarbon fractions to lighter, more valuable products such as light olefins, gasoline, and diesel. The catalyst used in FCC usually consists of FAU-type Y-zeolite, active matrix alumina, binder, and clay. In addition to the Y-zeolite based catalyst, the MFI-type ZSM-5 zeolite based catalyst, also referred to as an additive in FCC, is used to enhance the gasoline octane and boost propylene yield in the FCC unit. The demand for propylene is primarily driven by a rapid increase in polypropylene usage. In recent years, the global demand for propylene has expanded in line with the economic growth of newly developing countries such as China, India, and the Middle East. The trend toward the use of natural gas as a predominant feedstock in the steam crackers has shifted selectivity to ethylene production. Consequently, the supply of propylene to the global market is limited further. The demand for propylene is projected to increase by 5% each year.1 To keep up with demand, a significant increase in propylene production from FCC units is required. The FCC process consists of two coupled operations, feedstock cracking in the reactor and catalyst regeneration in the regenerator. Therefore, the fluidized catalyst is circulated between the riser reactor and the regenerator. In the bottom section of the riser reactor, the hot regenerated catalyst is contacted with the preheated heavy feedstock. The feed is vaporized and cracked within 1−3 s before the gaseous products are separated from the catalyst at the top of the riser. A typical reactor works with an outlet temperature of 540 °C, a feed throughput of 7 t min−1, and a catalyst-to-oil ratio of 6. The spent catalyst is then stripped with steam to remove any remaining hydrocarbons and subsequently reactivated in the regenerator, where the carbon deposited during the cracking reaction is burnt off with air at approximately 700 °C. The FCC © 2014 American Chemical Society

catalyst is hydrothermally deactivated because of the harsh conditions of the FCC process.2 Catalyst deactivation and losses by attrition of the catalyst particles are counterbalanced by daily fresh catalyst additions that result in an age distribution in the catalyst inventory. The fluidized catalyst inventory is called the equilibrium catalyst (Ecat).3 Zeolites are microporous, crystalline aluminosilicates that have an up to three-dimensional pore structure. The variation of the Si/Al ratio in a zeolite framework type changes its pore size, acidity, and catalytic properties. During the cracking process, the catalyst is deactivated because of the hydrothermal conditions in the regenerator that lead to the dealumination of the zeolite framework. For Y-zeolite, the change in the Si/Al ratio was correlated with its unit cell size (UCS).4 In the case of Y-zeolite, the UCS is defined as the length of the edge of the cubic unit cell. The UCS of Y-zeolite can be determined by powder X-ray diffraction (PXRD) with a precision of better than 0.01 Å from the peak position of an accurately measured reflection. To correct for any preparation and instrument errors, silicon powder with precisely known reflection positions can be added to the sample as an internal standard.5 For ZSM-5, which crystallizes with orthorhombic symmetry,6 the lengths of all three edges (a ≠ b ≠ c) are necessary to describe the unit cell volume (UCV). Consequently, the UCV cannot be determined by measuring the peak position of a single reflection. Also, the Si/Al ratio is significantly higher compared to that of Y-zeolite, which would result in a small shift of the peak position with further framework dealumination. In this case, the use of Rietveld refinement, which Received: Revised: Accepted: Published: 16270

June 23, 2014 September 19, 2014 September 23, 2014 September 23, 2014 dx.doi.org/10.1021/ie502522q | Ind. Eng. Chem. Res. 2014, 53, 16270−16274

Industrial & Engineering Chemistry Research

Article

were fixed after a pure deactivated ZSM-5 sample was measured. 2.3.1. Precision of the Method. Prior to discussion of the UCV results, it is vital to know the precision of the method to determine whether the observed differences in the UCV are significant or within the standard deviation. Therefore, a deactivated sample of a ZSM-5 containing olefins additive was prepared and measured by three different operators on four different days with the same instrumentation. The mean UCV measured for this sample was 5356.4 Å3 with a standard deviation of 0.47 Å3. However, preparation and measurement of the samples in one workflow by the same operator can decrease the standard deviation to 0.28 Å3. 2.4. Determination of Surface Area. The SA of the samples was determined by nitrogen sorption measurements using a Micromeritics Tristar 3000 and application of the Brunauer−Emmett−Teller (BET) theory.13 2.5. Acidity Measurements by Temperature-Programmed Desorption. The total acidity of the samples was determined by temperature-programmed desorption (TPD) of ammonia using a Thermo Finnigan TPDPRO 1100. Under a helium atmosphere, approximately 400 mg of sample was heated to 540 °C and allowed to cool to 150 °C. In a second step, the sample was saturated with ammonia at 150 °C, and then excess ammonia was flushed with helium while the sample was cooled to 60 °C. The actual ammonia desorption measurement took place by heating the sample with a ramp of 20 °C min−1 to 540 °C under a helium atmosphere, and by quantifying the released ammonia. 2.6. Testing Catalytic Properties. The catalytic activity and selectivity were measured by an advanced cracking evaluation (ACE) unit.14 Every sample was measured at six different catalyst-to-oil ratios that ranged from 3.5−7.5 at a temperature of 527 °C. Feed (1.5 g) (density of 0.9141 g cm−1; sulfur 0.4 wt %; Conradson carbon 1.6 wt %) was injected into the catalyst bed within 30 s. All of the results reported are interpolated data at a constant conversion of 70 wt % fresh feed (FF).

evaluates the whole PXRD pattern, is advantageous to get the necessary number of observations and sensitivity. This method uses an algorithm to achieve a best match of a calculated pattern with the observed data.7,8 The parameters of the calculated pattern contain all of the information necessary to derive the UCV of ZSM-5 with high precision. This paper shows the importance of the UCV in the characterization of ZSM-5. Our results show that the UCV of ZSM-5 in olefins additives can be measured accurately enough with laboratory PXRD equipment to monitor different stages of the ZSM-5 zeolite dealumination process. The data exhibit a good correlation between UCV and propylene yield, which allows a prediction of propylene yield from the FCC process. The data correlate very well with the other techniques, such as surface area (SA) and acidity measurement, that are used for the characterization of ZSM-5.

2. EXPERIMENTAL SECTION 2.1. Laboratory Deactivation Protocol. The catalysts and additives were deactivated prior to the study of their catalytic activity and properties. To simulate a field deactivation of the FCC process, fresh olefins additives containing ZSM-5 were deactivated in a quartz reactor with pure steam at 816 °C for 24 h, unless otherwise stated. The steam flow rate was adjusted to achieve a bubbling fluidized bed under steaming conditions.9 2.2. Sink−Float Density Separation Technique. The ZSM-5 containing additives were isolated from the FCC base catalyst by a sink−float density separation technique as described by Boock et al.10 The density separations were performed by adding the catalyst sample to a mixture of tetrabromoethane (TBE) and tetrachloroethane (TCE), which have density values of 2.96 g cm−3 and 1.58 g cm−3, respectively. The density of the olefins additive is lower than that of the FCC base catalyst. The density of the solvent was adjusted by controlling the TBE/TCE ratio. Separation of the catalyst components occurs when the density of the TBE/TCE mixture is in between the skeletal density value of the olefins additive and that of the base catalyst. The metal deposition on the catalyst from the feedstock also increases the skeletal density of the particles, which allows for further separation into age fractions. 2.3. Powder X-ray Diffraction and Rietveld Refinement. All of the catalyst and ZSM-5 containing additive samples were milled with a HERZOG HP-MS mill and spiked with ∼10 wt % silicon powder (Alfa Aesar, crystalline, 325 mesh, 99.5%), which was used as an internal standard to correct for preparation and instrument errors, and thus increase the accuracy of the method. The samples were measured on a Bruker D8 Advance diffractometer using Cu Kα radiation and a VÅNTEC-1 detector. Diffraction patterns were obtained from 3−70° with a step size of 0.008° 2θ. The total data collection time was 58 min per sample. The UCV was calculated from the resulting diffraction patterns using the Rietveld refinement function of the TOPAS software package.11 The phases that were refined with structure models are silicon (for angular correction), anatase, quartz, kaolinite, gamma-Al2O3, mullite, and AlPO4. The ZSM-5 phase itself was modeled by a Pawley12 fit, which takes only into account the symmetry information on the phase; however, to prevent the program from modeling undefined, overlapping reflections of ZSM-5, the relative intensities of the reflections,

3. RESULTS AND DISCUSSION 3.1. Laboratory Deactivated Additives. To study the dealumination process, a sample of commercially available ZSM-5 containing additive was deactivated at 816 °C for 24 h at various steam partial pressures ranging from 0−100% atmospheric pressure. The resulting samples were analyzed for SA, total acidity, and UCV. The SA is plotted against the steam partial pressure, and Figure 1, panel (a) shows the initial SA of the fresh sample to be at 149 m2 g−1 followed by an increase in SA to 156 m2 g−1 after mild steaming with 25% steam partial pressure, which dropped down to 138 m2 g−1 with 100% steam partial pressure. The initial low SA of the sample before deactivation is attributed to the blocking of the micropores by the binder, which opened up after the mild steaming. As shown in Figure 1, panel b, the initial total acidity of the fresh catalyst samples was reduced from 0.076 μmol NH3 m−2 to 0.024 μmol NH3 m−2 under mild steaming conditions with 25% steam partial pressure and further reduced to 0.016 μmol NH3 m−2 with 100% steam partial pressure. The UCV follows this trend by decreasing 6 Å3 from 5372 Å3 to 5366 Å3 with 25% steam partial pressure and decreasing further to 5360 Å3 with 100% steam partial pressure. The plot shows that the measured UCV correlates very well with the total acidity. In 16271

dx.doi.org/10.1021/ie502522q | Ind. Eng. Chem. Res. 2014, 53, 16270−16274

Industrial & Engineering Chemistry Research

Article

Figure 2. Delta propylene yield and UCV plotted against (a) the steaming duration at 816 °C and (b) the steaming temperatures for 24 h.

Figure 1. (a) SA and (b) UCV with total acidity plotted against the steam partial pressure. Samples were deactivated at 816 °C for 24 h at various steam concentrations. The data points at 0% steam represent the fresh catalyst sample before deactivation.

contrast to the SA, the UCV and total acidity show the changes of the ZSM-5 at mild steam pressures of 25%. These results prove that small differences in the lattice parameters can be verified using a modern laboratory PXRD with Rietveld analysis and that the UCV can be used as a measure of the ZSM-5 dealumination. To correlate the UCV with the performance of a commercially available ZSM-5 containing additive, a sample was deactivated in the laboratory at various steaming conditions. For the first set, the steaming time varied from 24−72 h at 816 °C, and for the second set, the temperatures varied from 816−871 °C at a constant steaming time of 24 h. To evaluate the propylene yields, the additive samples were blended at 5 wt % additive loading to a standard Ecat sample and tested in the ACE unit. The performance of a pure Ecat sample was used as the reference to calculate the delta propylene. The physical and catalytic properties of the deactivated additive are summarized in Table 1. Figure 2, panel (a) shows the plot of delta propylene and UCV were

Figure 3. Delta propylene yield plotted against the UCV of samples deactivated at various conditions.

plotted against steaming duration. While the delta propylene yield decreases by 0.7 wt % FF, the UCV decreases by 3 Å3. As pointed out in Figure 2, panel b, the steaming temperature has a strong effect on both the propylene yield and the UCV, which

Table 1. Physical and Catalytic Properties of the Commercially Available ZSM-5 Containing Additive after Deactivation at Various Steaming Durations and Temperatures steaming duration (h) 24 48 72 24 24

steaming temperature (°C) 816 816 816 843 871

SA (m2 g−1)

UCV (Å3)

delta proplylene yield (wt % FF)

148 144 136 132 109

5359 5358 5356 5357 5354

2.6 2.0 1.9 2.0 1.5

16272

dx.doi.org/10.1021/ie502522q | Ind. Eng. Chem. Res. 2014, 53, 16270−16274

Industrial & Engineering Chemistry Research

Article

Table 2. Physical and Catalytic Properties of ZSM-5 Additive Separated from Ecat and Further Separated into Different Age Fractions fraction

Ni+V (mg kg−1)

SA (m2 g−1)

UCV (Å3)

delta propylene yield (wt % FF)

youngest fraction second youngest fraction second oldest fraction oldest fraction

2526 3508 4910 6892

149 145 143 138

5365 5360 5359 5356

4.6 3.7 3.2 1.8



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Author Contributions

The manuscript was written through the contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■ Figure 4. Delta propylene yield plotted against the UCV of the samples separated from Ecat into age fractions.

decrease from 2.6−1.5 wt % FF and 5359−5354 Å3, respectively. Figure 3 plots a linear correlation between the delta propylene yield and the UCV, which allows a prediction of the ZSM-5 additive performance. 3.2. Commercially Deactivated Additives. The ZSM-5 containing additive of a commercial Ecat sample was separated from the base catalyst by the sink−float procedure. The resulting ZSM-5 additive was further separated into age fractions based on its density. Table 2 summarizes the chemical and physical properties of the different age fractions separated from the Ecat by the sink−float procedure and the catalytic properties after blending at 5 wt % with a standard Ecat. The increase in the metal concentration (Ni+V) is a good indicator of the age of the fractions separated by a sink−float method; however, in contrast to Y-zeolite, it is not expected that the metals will have dealumination effects on the ZSM-5.15 The delta propylene yield was plotted against the UCV of the different fractions and is shown in Figure 4, which again shows a clear positive correlation between the ZSM-5 activity and the UCV.



ABBREVIATIONS ACE = advanced cracking evaluation BET = Brunauer−Emmett−Teller FCC = fluid catalytic cracking PXRD = powder X-ray diffraction SA = surface area TBE = tetrabromoethane TCE = tetrachloroethane TPD = temperature-programmed desorption UCS = unit cell size (of Y-zeolite) UCV = unit cell volume (of ZSM-5 zeolite) REFERENCES

(1) Hamada, R.; Watabe, M. 18th Annual Saudi−Japan Symposium on Catalysts in Petroleum Refining & Petrochemicals; King Fahd University of Petroleum and Minerals: Dhahran, Saudi Arabia, 2008. (2) Mante, O. D.; Agblevor, F. A.; Oyama, S. T.; McClung, R. The effect of hydrothermal treatment of FCC catalysts and ZSM-5 additives in catalytic conversion of biomass. Appl. Catal., A 2012, 445− 446, 312. (3) Cheng, W.-C.; Habib Jr., E. T.; Rajagopalan, K.; Roberie, T. G.; Wormsbecher, R. F.; Ziebarth, M. S. Fluid Catalytic Cracking. In Handbook of Heterogeneous Catalysis; Ertl, G., Knözinger, H., Schüth, F., Weitkamp, J., Eds.;Wiley-VCH: Weinheim, Germany, 2008. (4) Breck, D. W.; Flaningen, E. M. Synthesis and properties of unioncarbide zeolites L, X, and Y. Mol. Sieves, Soc. Chem. Ind. 1968, 47−61. (5) Method D3942−08. Standard Test Method for Determination of the Unit Cell Dimension of a Faujasite-Type Zeolite; ASTM International: West Conshohocken, PA, 2008. (6) Olson, D. H.; Kokotailo, G. T.; Lawton, S. L.; Meier, W. M. Crystal structure and structure-related properties of ZSM-5. J. Phys. Chem. 1981, 85, 2238. (7) Rietveld, H. M. Line profiles of neutron powder-diffraction peaks for structure refinement. Acta Crystallogr. 1967, 22, 151. (8) Rietveld, H. M. A profile refinement method for nuclear and magnetic structures. J. Appl. Crystallogr. 1969, 2, 65. (9) Method D4463/D4463M−96. Standard Guide for Metals Free Steam Deactivation of Fresh Fluid Cracking Catalysts; ASTM International: West Conshohocken, PA, 2012. (10) Boock, L. T.; Deady, J.; Lim, T. F.; Yaluris, G. New developments in FCC catalyst deactivation by metals: Metals mobility and the vanadium mobility index (VMI). In Studies in Surface Science and Catalysis: Catalyst Deactivation; Elsevier: Amsterdam, The Netherlands, 1997.

4. CONCLUSION In this paper, we have shown the significance of UCV determination for the characterization of ZSM-5 in the additive formulation. It was demonstrated that the technique is valid for laboratory deactivated and commercially deactivated ZSM-5 additives. For Y-zeolite, measurement of the UCS is a widely used method to monitor deactivation and predict the catalytic activity and selectivity of a catalyst. This concept was transferred to ZSM-5 zeolite, where the UCV is used instead. Modern laboratory PXRD equipment and Rietveld analysis allows the measurement of the UCV with sufficient sensitivity to monitor the dealumination of ZSM-5 in a laboratory and commercially deactivated catalyst. This method can be used further to understand the fundamentals of the deactivation and hence to improve the stability and activity of ZSM-5 in the catalyst for future applications. 16273

dx.doi.org/10.1021/ie502522q | Ind. Eng. Chem. Res. 2014, 53, 16270−16274

Industrial & Engineering Chemistry Research

Article

(11) Cheary, R. W.; Coelho, A. A fundamental parameters approach to X-ray line-profile fitting. J. Appl. Crystallogr. 1992, 25, 109. (12) Pawley, G. S. Unit-cell refinement from powder diffraction scans. J. Appl. Crystallogr. 1981, 14, 357. (13) Brunauer, S.; Emmett, P. H.; Teller, E. Adsorption of gases in multimolecular layers. J. Am. Chem. Soc. 1938, 60, 309. (14) Kayser, J. C. Versatile Fluidized Bed Reactor. U.S. Patent 6,069,012, May 23, 1997. (15) Rautiainen, E.; Pimenta, R.; Ludvig, M.; Pouwels, C. Deactivation of ZSM-5 additives in laboratory for realistic testing. Catal. Today 2009, 140, 179.

16274

dx.doi.org/10.1021/ie502522q | Ind. Eng. Chem. Res. 2014, 53, 16270−16274