How Feed, Support, and Reductant Drive Coking in Pt–Sn Dehydrogenation Catalysts

The world’s plastics, fuels, and countless everyday chemicals begin with a deceptively simple chemical transformation: stripping hydrogen from saturated hydrocarbon molecules known as paraffins to produce olefins, the reactive building blocks of modern industry. Yet the catalysts that drive this reaction, typically platinum particles promoted with tin and dispersed on alumina supports, have long been plagued by an unavoidable nemesis known as coke, carbon-rich deposits that accumulate on active sites and slowly choke off catalytic activity. A new study published in Catalysis Letters by Suresh Avithi Kanniappan and Udaya Bhaskar Reddy Ragula of Amrita Vishwa Vidyapeetham in Coimbatore, India, offers one of the most detailed quantitative pictures to date of how these catalysts die, how the carbonaceous deposits that kill them form and evolve, and, critically, how quickly and by what mechanism they can be brought back to life during regeneration.

The researchers synthesized six different bimetallic platinum-tin catalysts using a wetness impregnation method, deliberately varying two key parameters: the chemical identity of the reducing agent used to activate the metal precursors, and the crystalline phase of the alumina support on which the metals were dispersed. This systematic approach allowed the team to disentangle effects that are often conflated in industrial practice. Each catalyst was then subjected to dehydrogenation reactions in a fixed-bed reactor, both with individual paraffin feeds and with mixed-paraffin feeds, a strategy the authors had previously proposed as a route to higher overall conversion without sacrificing the selectivity to valuable olefins that makes platinum-tin systems so commercially attractive.

The central technical achievement of the work lies in its quantitative treatment of deactivation. By recording catalytic activity as a function of time-on-stream and fitting the resulting decay curves with kinetic models, the team extracted deactivation rate constants for every catalyst-feed combination. The analysis revealed a striking uniformity: regardless of the support, the reducing agent, or the composition of the paraffin feed, all six catalysts followed second-order deactivation kinetics. This means the rate at which activity is lost is proportional to the square of the remaining activity, a mathematical signature suggesting that carbon deposition accelerates as the catalyst degrades, perhaps because partially coked surfaces promote further coke formation or because the loss of active sites concentrates reactions on fewer, more heavily burdened locations.

To characterize the coke itself, the team turned to thermogravimetric analysis, burning off the deposits under a zero-air atmosphere and carefully recording the mass loss as temperature increased. The differential thermogravimetric data revealed not a single combustion event but two distinct peaks, indicating that the coke on these spent catalysts exists in at least two populations with different chemical identities and thermal stabilities. Using the Fraser-Suzuki function, a flexible asymmetric peak-shape model widely used to deconvolute overlapping thermal events, the researchers mathematically separated these two contributions and then applied the Coats-Redfern method to each isolated peak independently to extract kinetic parameters for the coke removal process.

The results of that deconvolution are particularly valuable for catalyst regeneration engineering. The lower-temperature peak was best described by a random-scission model, the same kinetic framework used to describe the thermal degradation of polymers, suggesting a less ordered, more aliphatic or hydrogen-rich carbon material that breaks apart in a statistically random fashion as it oxidizes. The higher-temperature peak, by contrast, obeyed a second-order kinetic model, consistent with a more condensed, graphitic carbon structure whose combustion is limited by the interaction of reacting carbon domains. Spectroscopic confirmation came from Fourier-transform infrared and Raman spectroscopy, which probed the functional groups and degree of structural order in the deposits, corroborating the coexistence of distinct coke species on the spent catalysts.

Perhaps the most consequential single number to emerge from the study is the range of activation energies for the decoking process: between 150 and 216 kilojoules per mole. These values, spanning the two combustion regimes and the six catalyst formulations, quantify the thermal energy barrier that must be overcome to burn the coke away. In practical terms, this range tells process engineers how sensitive the regeneration step is to temperature and provides a direct input for designing industrial regeneration cycles that remove coke completely without exposing the expensive platinum metal to sintering, the process by which metal particles coalesce at high temperatures and permanently destroy active surface area.

The industrial stakes of this work are considerable. Light olefins such as propylene and isobutene are among the highest-demand commodity chemicals on the planet, serving as feedstocks for polypropylene, acrylonitrile, and a vast range of polymers and intermediates. Catalytic dehydrogenation of paraffins in packed-bed reactors is a major route to these molecules, and the dominant commercial technologies rely on platinum-tin catalysts celebrated for their exceptional olefin selectivity even at low platinum loadings. The persistent weakness has always been conversion: thermodynamic and kinetic constraints keep single-paraffin conversion modest, while coking forces repeated shutdowns for regeneration. Mixed-paraffin dehydrogenation, as explored here, offers a way to push conversion higher while keeping selectivity essentially unchanged, but it also changes the coking environment in ways that must be understood before commercial adoption.

This is where the study’s comparison of feeds, supports, and reducing agents becomes more than an academic exercise. The nature of the feed determines both the amount and the character of the coke deposited; heavier or more branched paraffins tend to generate different carbonaceous residues than lighter straight-chain molecules. The alumina support phase governs metal dispersion, acidity, and the strength of metal-support interactions, all of which influence where coke forms, whether on the metal particles themselves, where it poisons dehydrogenation sites directly, or on the acidic support surface, where it may block pores and alter transport. The reducing agent, meanwhile, shapes the alloying state of the platinum-tin bimetallic surface, which is known to be central to both activity and coke resistance, since tin dilutes ensembles of platinum atoms and moderates their tendency to catalyze dehydrogenative carbon formation.

The choice of spectroscopic tools reflects the study’s dual focus on quantity and quality of coke. Raman spectroscopy, with its characteristic D and G bands, provides a rapid readout of the degree of graphitization in carbon deposits, distinguishing disordered, reactive carbon from condensed, refractory material. FTIR spectroscopy reveals oxygen-containing and aliphatic functional groups that mark the early stages of coke growth. By combining these molecular-level fingerprints with the bulk thermal analysis and the formal kinetic modeling, the authors constructed a coherent picture linking coke chemistry to combustion behavior: the aliphatic, polymer-like fraction burns readily through random-scission-type kinetics, while the condensed aromatic fraction demands more energy and follows second-order combustion dynamics.

For the field of catalytic process engineering, the significance of this study lies in its completeness. Deactivation kinetics tell reactor designers how catalyst performance will decline over a cycle and therefore how frequently beds must be switched or regenerated. Decoking kinetics and activation energies tell them how the regeneration step should be run: at what temperature, for how long, and with what sensitivity to heat-up rates. The identification of two coke populations with distinct combustion models suggests that regeneration protocols could be optimized to first remove the labile fraction at moderate temperatures and then address the refractory fraction, potentially allowing partial regenerations that extend catalyst life while minimizing thermal stress. The finding that second-order deactivation holds universally across all six catalysts simplifies reactor modeling, offering a single robust functional form for use in packed-bed simulations of industrial dehydrogenation units.

The work also builds directly on the authors’ own earlier research, which examined how different alumina supports and reduction methods affect platinum-tin catalyst performance in individual and mixed-paraffin dehydrogenation, and which explored the effect of catalyst reduction state on mixed-paraffin feed processing. The new study extends that program from performance characterization to the fundamentally important question of catalyst death and rebirth, closing the loop between catalyst formulation, operational behavior, and regenerability. Funded by India’s Science and Engineering Research Board and the Council for Scientific and Industrial Research, the research exemplifies the kind of systematic, quantitative catalysis work that underpins incremental but essential improvements in one of the chemical industry’s most energy-intensive and capital-heavy processes. As global demand for light olefins continues to climb, understanding the precise kinetics of how catalysts fail and how they recover may prove as valuable as any advance in making them more active in the first place.