Research ArticleSURFACE CHEMISTRY

Life and death of a single catalytic cracking particle

See allHide authors and affiliations

Science Advances  03 Apr 2015:
Vol. 1, no. 3, e1400199
DOI: 10.1126/sciadv.1400199
  • Fig. 1 Bulk analysis results of FRESH, LML, MML, and HML FCC catalyst particles.

    (Top) Fresh catalyst particles and calcined ECat particles were obtained from an industrial FCC unit and separated into three age groups according to their skeletal density (ρ1 to ρ3). (Middle) Catalytic cracking activity (%) (blue) of vacuum gas oil (VGO) and accessibility index (purple) are anticorrelated with the total metal content of Fe and Ni (summed Fe and Ni concentrations in wt %) for particle surface (green) and bulk (red), as measured by SEM-EDX and WDXRF (Table 1). (Bottom) Catalytic particles are imaged by TXM tomography at 64-nm 3D voxel size (see the Supplementary Materials). Red to orange colors indicate Fe concentrations, whereas Ni is visualized using the blue to green color range (see also movie S1).

  • Fig. 2 Transmission x-ray nanotomography of FCC catalyst particles.

    (A and B) Data are collected below and above the x-ray absorption Fe and Ni K-edge, respectively (A), and reconstructed separately, resulting in four sample volumes (B). (C and D) Pairwise subtraction of the volumes (differential absorption contrast imaging) provides the 3D distribution of Fe (red) and Ni (green) (C), which is then correlated with pore distribution analysis to reveal the degree of pore clogging by metals and the effects on pore connectivity and accessibility (D). Massive data sets collected and analyzed for every single FCC particle contained, depending on the individual particle size, between 301 and 836 million voxels of 64 × 64 × 64 nm3. See movie S1 displaying a fully reconstructed particle illustrating Fe and Ni distributions.

  • Fig. 3 Radial evaluation of single FCC particles.

    (A to D) Elemental concentrations of Fe (A) and Ni (C) and porosity changes caused by Fe (B) and Ni (D) in FCC particles of different ages (HML, MML, LML, and FRESH), plotted as a function of distance from the particle surface. For the FRESH particle, only Fe was analyzed because it contained no Ni. Data analysis methods and evaluation of uncertainty used to generate fan plots are reported in the Supplementary Materials.

  • Fig. 4 Changes in the pore accessibility of a single FCC catalyst particle as a function of metal content and type.

    (Top) Scheme for metal deposition in the pore network, indicating Fe (red) mainly at the outer surface and Ni (green) penetrating more deeply, as demonstrated in Fig. 3. Pore channels can remain open (a), become narrowed (b), or have blocked accessibility (c) by metal deposition in surface pore access sites. Case “c” may also be that macropores (>64 nm) have been converted to mesopores (<64 nm) that are no longer visible with this method. (Middle) Pore radius change in percent as a function of particle age for Fe (red) and Ni (green), respectively. For the FRESH sample, the pore radius change is negligible, indicating that the pore network is not affected by macropore narrowing (that is, case “a” represented by the corresponding inset). The two other insets schematically display how pore narrowing becomes significant for the LML and MML samples and is dominated by Ni in the early phase (LML sample). (Bottom) Decrease of surface access points in percent as a function of particle age for Fe (red) and Ni (green), respectively. The two insets schematically depict how at an earlier stage (LML) Ni dominates the reduction of surface macropores that provide access to the internal pore network of the particle (case c), while later (MML and HML), both metals cause severe access restriction through macroporosity.

  • Fig. 5 Effects of Fe and Ni on pore network interconnectivity and accessibility from the surface.

    (Left) Rendering of TXM tomography data collected for the FRESH sample including 3D Fe concentrations (top panel) and the HML sample without metals (second panel), including Fe distribution (third panel, same dimensions), and both Fe and Ni (bottom panel, same dimensions). (Middle) 3D representation of the established pore networks; for clarity, only the main network is plotted for the FRESH and HML samples in the top and second panel, and the largest 5 and 10 pore networks in terms of volume are plotted in panels 3 and 4 from the top, respectively. (Right) Mollweide projections of the particle surface depicting all detected access sites connected to different macropore networks in the particle. Green markers provide access to the largest interconnected macropore subvolume (given in percent of the TPV). For the HML particle, Ni completely blocks surface access to the largest macropore subvolume. Access to smaller macropore subvolumes is indicated by the color scale ranging from 0 to 2% TPV. Gray pixels (0%) refer to areas without surface access points, that is, without detectable macropores. See the Supplementary Materials for analytical details and Fig. 6 for similar maps of correlated penetration depths.

  • Fig. 6 2D surface maps visualizing access points to the macropore volume of single FCC catalyst particles—penetration depth.

    Effect of metal poisoning by Fe (left) and Ni (right) restricting access to the FCC particle pore volume, visualized as a 2D surface map (Mollweide projection). Dots indicate remaining access sites after considering the access-blocking effect of the metals. Green markers provide access to the largest interconnected macropore subvolume and therefore allow the entering feedstock molecules to reach the largest penetration depth (given in percent of the particles’ equivalent spherical radii, see table S1). For the HML particle, Ni completely blocks surface access to the largest macropore subvolume. The color scale ranging from 0 to 35% of the equivalent spherical particle radius indicates access to smaller macropore subvolumes. Gray pixels (0%) refer to areas without surface access points, that is, without detectable macropores.

  • Table 1 Results of bulk (multiple particle) analysis performed with the density-separated particle fractions.

    Uncertainties are reported as 1σ and have been estimated from repeated control measurements where possible [unknown SDs are indicated by N/K (not known)]. Nickel and vanadium were not detected (N/D) in SEM-EDX and WDXRF analyses of the FRESH sample. As explained in the text, the conversion of the FRESH sample is assumed to approach 100 wt %.

    MetricFRESHLMLMMLHML
    Skeletal density (g/cm3)2.7852.9462.9532.957N/K
    SEM-EDX, La (wt %)1.641.701.631.84N/K
    SEM-EDX, Fe (wt %)0.271.341.601.61N/K
    SEM-EDX, Ni (wt %)N/D0.340.350.40N/K
    SEM-EDX, V (wt %)N/D0.200.220.29N/K
    WDXRF, La2O3 (wt %)2.862.812.762.740.02
    WDXRF, Fe2O3 (wt %)0.340.790.930.960.01
    WDXRF, NiO (wt %)N/D0.330.490.590.01
    WDXRF, V2O5 (wt %)N/D0.530.660.740.01
    BET surface area (m2/g)262.0133.9108.193.32.62
    Micropore surface area (m2/g)131.581.673.062.53.66
    Micropore volume (cm3/g)0.06050.02430.01630.0140N/K
    Mesopore surface area (m2/g)139.052.335.130.83.73
    (Pore) accessibility (AAI)94.93.52.40.90
    430°F+ conversion (wt %) [CTO: 3 (w/w)]~10065.1257.5954.520.57
    430°F+ conversion (wt %) [CTO: 4 (w/w)]~10068.1661.2958.20.57
    430°F+ conversion (wt %) [(CTO: 5 (w/w)]~10070.9165.6460.930.57
    430°F+ conversion (wt %) [CTO: 6 (w/w)]~10074.4467.4664.210.57
  • Table 2 Changes in the single-particle macropore network caused by the metals Fe and Ni.

    Uncertainties have been determined as the SD of the changes (in percent) calculated using methods 1 and 2 for pore volume determination (see the Supplementary Materials). Nickel was not detected in the FRESH sample; therefore data on changes caused by Ni are not available (N/A).

    Pore volumeMetricFRESHLMLMMLHML
    FeChange in average pore radius (%)2.0 ± 0.98.8 ± 4.013.3 ± 4.66.8 ± 2.6
    Decrease in number of surface access points of largest subnetwork (%)17.2 ± 9.439.3 ± 6.794.5 ± 1.693.7 ± 0.6
    NiChange in average pore radius (%)N/A21.5 ± 10.520.0 ± 2.113.9 ± 3.4
    Decrease in number of surface access points of largest subnetwork (%)N/A80.9 ± 3.184.9 ± 0.9100.0 ± 0.0

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/1/3/e1400199/DC1

    Materials and Methods

    Fig. S1. Schematics of the topological representation of the pore network.

    Fig. S2. FRC for estimate of 3D resolution.

    Table S1. Single-particle metrics from TXM tomography data.

    Table S2. Basic parameters of the established macropore network.

    Movie S1. Fe and Ni distribution on and in a single MML catalyst particle.

    Movie S2. Visualizing the changes to the pore network with the presence of Fe and Ni in the pores.

    References (46, 47)

  • Supplementary Materials

    This PDF file includes:

    • Materials and Methods
    • Fig. S1. Schematics of the topological representation of the pore network.
    • Fig. S2. FRC for estimate of 3D resolution.
    • Table S1. Single-particle metrics from TXM tomography data.
    • Table S2. Basic parameters of the established macropore network.
    • Legends for movies S1 and S2
    • References (46, 47)

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • Movie S1 (.mov format). Fe and Ni distribution on and in a single MML catalyst particle.
    • Movie S2 (.mov format). Visualizing the changes to the pore network with the presence of Fe and Ni in the pores.

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article