Molecular elucidating of an unusual growth mechanism for polycyclic aromatic hydrocarbons in confined space

The study addresses how polycyclic aromatic hydrocarbons (PAHs) form, grow, and cluster under confined conditions in zeolite- or molecular sieve-catalyzed reactions—processes central to catalyst coking and deactivation in industrial petrochemical and coal-based conversions such as the methanol-to-olefins (MTO) process. While lighter, soluble PAHs have been partially characterized, the molecular structures and evolution of heavier, insoluble PAHs remain elusive due to analytical limitations. The authors hypothesize that zeolite confinement dictates PAH size and structure and seek to reveal a complete molecular-level route of PAH growth and clustering inside SAPO-34 cages under high-temperature, industrially relevant MTO conditions, with particular focus on identifying any previously unrecognized growth steps.

Conventional analysis of coke species combines HF dissolution, solvent extraction, and GC-MS (Guisnet’s method), but is limited to molecules <300 g mol−1 due to solubility constraints. Spectroscopic techniques (IR, UV-Raman, UV-Vis, and 13C NMR) can distinguish soluble vs. insoluble fractions and be used in situ/operando but cannot provide detailed molecular structures of insoluble, highly condensed PAHs. Microscopic methods (confocal fluorescence microscopy, atom probe tomography) have elucidated spatial distributions of coke and relations to acid sites but lack molecular structural fingerprints. Mass spectrometry approaches (tandem MS, HPLC/UV/MS, synchrotron VUV photoionization MS) have been applied to combustion- and pyrolysis-derived PAHs but struggle with isomer resolution, often requiring standards. MALDI FT-ICR MS has emerged as a soft-ionization tool for coke structural identification in zeolite-catalyzed reactions (e.g., HBEA, HMOR, HZSM-5), yet alone is insufficient to resolve isomeric complexity. These gaps motivate an integrated strategy combining MALDI FT-ICR MS with isotopic labeling and complementary spectroscopic and computational tools.

- Model reaction and reactor configuration: SAPO-34-catalyzed MTO carried out in a fixed-bed quartz reactor (8 mm i.d.), atmospheric pressure. Catalyst (600 mg, 250–420 µm) was divided into four equal layers by quartz wool to reduce axial inhomogeneity. Activation at 480 °C in N2 for 1 h; typical reaction at 475 °C with WHSV = 4 h−1; methanol fed via N2 through a 25 °C saturator. Effluent analyzed online by GC (HP-Plot/Q-HT, FID); conversions/selectivities on CH2 basis. - Catalysts: SAPO-34 (CHA), SAPO-18 (AEI), SAPO-35 (LEV), DNL-6 (RHO) synthesized hydrothermally with specified organic structure-directing agents; calcined to H-form. Physicochemical properties measured (XRD, SEM, XRF, N2 sorption t-plot/BET, XPS, 1H/29Si MAS NMR). - Coke liberation and fractionation: Modified HF dissolution–CCl4 extraction liberates retained organics from spent catalysts. Improvements over classical method: no HF neutralization (avoids heat/AlF3 formation) and use of denser, less water-soluble CCl4 as extractant to enhance phase separation and recovery. Soluble fraction analyzed by GC-MS with C2Cl6 internal standard and calibration; insoluble fraction isolated by suction filtration and CCl4 washing. - Mass spectrometry: MALDI FT-ICR MS (15 T Solarix XR, Nd:YAG 355 nm, positive ion mode, 150 < m/z < 1200; 200 Hz; 32 shots) using dithranol matrix in THF. Three modes: (i) direct analysis of deactivated catalysts to check external-surface coke; (ii) analysis of CCl4 extracts; (iii) analysis of isolated insoluble coke. Control experiments determined optimal laser output (18%) to avoid polymerization/cleavage. Isotope labeling with 13C-methanol and deuterated methanol (CD3OD) to determine precise C and H counts of PAHs via mass shifts. - Solid-state NMR: 13C MAS NMR on isolated 13C-labeled coke (9.4 T, 0.7 mm H-X probe, 110 kHz MAS, HP decoupling); peak fitting to quantify aromaticity, substitution degree, and bridgehead carbon fraction. - Computational studies: DFT (DMol3) assessed adsorption energies and size feasibility of PAHs (up to pyrene) in single CHA cages. Periodic DFT (CP2K, PBE-D3, GTH-DZVP, plane-wave cutoff 650 Ry) on 2×2×1 AlPO-34 supercells optimized configurations and interactions of proposed multi-core PAHs spanning adjacent cages. - Additional analyses: TGA for total coke content; in situ/operando insights referenced for lattice expansion; data treatment included correlating micropore volume loss to coke content and comparing top vs. bottom reactor layers.

- Full-spectrum coke analysis: Combining GC-MS with MALDI FT-ICR MS extended detectable masses from <300 Da to 300–1200 Da, enabling molecular characterization of heavier, insoluble PAHs. - In-pore dominance and external surface exclusion: Multiple lines of evidence (product slate, low external area, linear micropore volume loss vs. coke content, direct MALDI on intact catalysts showing no significant external coke signals) indicate coke predominantly resides within SAPO-34 cages. - Size limit in single cages: DFT and GC-MS concur that pyrene and derivatives represent the maximum-sized aromatics stably accommodated in a single CHA cage; larger five- to six-ring condensed aromatics are energetically unfavorable (positive adsorption energies >110 kJ mol−1). - Discrete MALDI mass groups: Insoluble PAHs show Gaussian mass groups with CH2 (14 Da) increments and maxima near m/z 339, 537, 700, with regular inter-group intervals of ~164–214 Da corresponding to 3–4 ring units (e.g., fluorene 166, phenanthrene 178, pyrene 202). This piecewise distribution suggests stepwise addition of 3–4-ring units. - Cage-passing growth mechanism: Data support a previously unrecognized growth step where PAHs span adjacent CHA cages via 8-membered-ring windows, progressively cross-linking 3–4-ring building units into multicore structures. - Isotope labeling establishes stoichiometries: 13C- and D-methanol feeding yielded precise compositions for prominent peaks: C27H16, C43H22, C56H28. Series follow general formulas CnH2n−40 (27≤n≤31), CnH2n−66 (43≤n≤48), CnH2n−84 (56≤n≤61). Degree of unsaturation increases by 20, 33, and 43 across groups, consistent with ring condensation rather than stepwise unsaturation increases characteristic of channel zeolites. - Structural models and graphene-like nature: DFT-optimized models indicate biphenyl- and/or methylene-bridged multicore PAHs forming single-layer graphene-like 3D configurations spanning 2–4 cages. 13C MAS NMR shows high aromaticity (f ≈ 95.9%), low substitution degree (o ≈ 0.037), and bridgehead carbon fraction x ≈ 0.36 (between phenanthrene 0.286 and pyrene 0.375), implying average 3–4 rings per primary cluster within individual cages. - Evolution pathway in MTO: PAHs evolve from occluded long-chain olefins and polymethylbenzenes (seeding) to 3–4-ring PAHs (growing units), and finally to cross-linked multicore clusters via cage-passing (clustering). Microporous surface area and volume drop ~80% during pseudo steady-state, aligning with rapid in-pore coke buildup. - Generality across cage molecular sieves: In SAPO-35 (LEV, 7.3×6.3 Å), naphthalene/fluorene units couple via biphenyl bridges; in SAPO-18 (AEI, 12.7×11.6 Å), pyrene acts as the principal unit; in DNL-6 (RHO, 11.4×11.4 Å), cage-passing is impeded by double 8-MR separations (~180 pm), yielding coronene-type coke without cage-crossing—validating topology-controlled mechanisms. - Deactivation model: Si-rich near-surface zones lead to higher acid density and preferential formation of bulky, cross-linked multicore clusters in shell layers, impeding diffusion, restricting further in-core aromatic growth, and leaving interior acid sites underutilized.

The integrated analytical strategy resolves the long-standing challenge of molecularly identifying heavy, insoluble PAHs responsible for coking in confined catalytic environments. By showing that discrete mass groupings correspond to repeated 3–4-ring building units and demonstrating precise C/H stoichiometries via isotopic shifts, the study establishes a cage-passing condensation mechanism unique to cage-based frameworks like SAPO-34. This mechanism reconciles the apparent paradox of bulky PAHs forming within nanosized cages by revealing cross-linking across adjacent cages, and it explains observed deactivation patterns—rapid micropore loss and near-surface coke accumulation that hampers diffusion and catalyst utilization. The strong agreement among MALDI FT-ICR MS, isotope labeling, DFT energetics/geometry, and 13C ssNMR provides convergent validation. Extending the analysis to SAPO-35, SAPO-18, and DNL-6 demonstrates that cage connectivity (single vs. double 8-MR) governs whether cage-passing growth can occur, underscoring the role of host topology in dictating coke structure and, thus, deactivation behavior. These insights inform industrial strategies such as pre-coking and partial regeneration to steer selectivity (e.g., ethene) and mitigate deactivation in MTO, and they offer a unifying motif for multicore PAH formation relevant to combustion soot, asphaltenes, and coal.

This work introduces a versatile, high-resolution approach—MALDI FT-ICR MS integrated with isotope labeling, complemented by GC-MS, 13C ssNMR, and DFT—to decipher the molecular structures and evolution of heavy, insoluble PAHs in confined catalytic environments. It uncovers a previously unrecognized cage-passing growth mechanism in SAPO-34-catalyzed MTO, wherein 3–4-ring aromatic building units cross-link across adjacent cages to form graphene-like multicore PAHs, and validates the mechanism’s dependence on cage connectivity across different cage-structured sieves. The findings yield a mechanistic deactivation model linking near-surface multicore cluster formation to diffusion limitations and unutilized interior sites, offering guidance for process strategies (pre-coking, partial regeneration) to manage selectivity and catalyst lifetime. Future work could generalize this analytic framework to other catalytic systems and operating conditions, expand isomer-specific structural assignments (e.g., via tandem MS or additional spectroscopies), quantify kinetics of cage-crossing growth, and design frameworks that modulate cage connectivity to control coke architecture.

- Structural assignment ambiguity: While isotopic stoichiometries (CxHy) and mass intervals constrain possibilities, multiple isomers can fit; proposed structures are representative rather than unique. - Detection window: MALDI FT-ICR MS has ionization/detection limits; even heavier or more cross-linked species may remain undetected. - External coke not entirely excluded: Evidence indicates minimal external coking on SAPO-34, but trace amounts could be present. - Specific conditions and materials: Findings are derived under particular MTO conditions (e.g., 475 °C, WHSV 4 h−1) and selected cage-based sieves; behavior may vary with different feeds, temperatures, or framework chemistries. - Indirect spatial inference: Spatial distribution is inferred from combined methods and literature microscopy; direct molecular imaging of multicore structures within cages remains challenging.