Carbon black and hydrogen production from methane pyrolysis: measured and modeled insights from integrated gas and particle diagnostics in shock tubes

Abstract

Methane (CH4) pyrolysis is a promising route to co-produce hydrogen (H2) and carbon black (CB) while avoiding emissions associated with steam-methane reforming and furnace black processes. Model development of pyrolytic CB synthesis requires experimental observations of concurrent gas chemistry, particulate formation, and morphology. This work presents a combined experimental and modeling study of CH4 pyrolysis behind reflected shock waves in 5% CH4/Argon mixtures at post-reflected shock temperatures (T5) of 1850-2450 K and P5 around 4.5 atm. Laser absorption diagnostics quantified CH4, C2H4, and C2H2 mole fractions, while multiwavelength extinction (633 and 1064 nm) resolved time-dependent particle formation and the temperature-dependent evolution of optical maturity. Simulations reproduce small-molecule speciation well, but large variations in predicted polycyclic aromatic hydrocarbons (PAHs) persist among models. Coupled gas-particle simulations capture accurate volume fraction (fv) trends and the influence of gas dynamics but underpredict induction times at high T5. Samples collected at the shock tube endwall were analyzed by transmission electron microscopy (TEM) to quantify primary particle size distributions and nanostructure arrangement. Image segmentation and manual measurements showed reduced primary particle size growth (dp) with increasing T5, while graphitic nanostructure generally increased. This study provides an integrated benchmark for improving models of CB and H2 production from CH4 pyrolysis by constraining gas-phase kinetics, PAH-driven inception, particle dynamics, and particle maturity. The results highlight that accurate partitioning of mass between particle number and particle size is an important constraint for further model development.

Figures (27)

• Figure 1: Schematic of gas speciation, particle formation, and particle sampling diagnostic integration for shock-tube pyrolysis experiments.
• Figure 2: Segmentation of a representative image from 2132 K, 4.2 atm using Cellpose-SAM pachitariu_CellposeSAMS and the effective diameter metrics employed in this study: (a) Raw TEM image, which was cropped to the dark region for automated analysis, with the example contour (339) highlighted in red, (b) Processed TEM segmentation mask showing 678 detected particle contours, (c) Contour 339 and its effective primary particle diameter described by six different size metrics.
• Figure 3: Workflow for quantifying carbon nanostructure fringes from bright-field HRTEM images. (a) Raw micrograph. (b) Selected ROI after preprocessing. (c) Extracted lattice fringes overlaid on the original image (red). (d) Parallel, equal-length fringe pairs within individual stacks (green) used to estimate inter-fringe spacing.
• Figure 4: $P(t)$, $T(t)$, gas chemistry and carbon particle formation in CH$_4$ pyrolysis at 1984 K and 4.3 atm: (a) $P(t)$ (black) is relatively flat for the first 2 ms, a reflection off of the contact surface drives a rise between 2.5 and 3.5 ms before dropping in the expansion fan. Modeled $T(t)$ (green markers) drops due to endothermic initiation chemistry then follows $P(t)$ isentropically; (b) the reflected shock arrives at time-zero. Measurements of CH$_4$ conversion into C$_2$H$_4$ and C$_2$H$_2$ (thick lines) are made for the first 1 ms (free of particle interference) and extinction at 633 nm (red) and 1064 nm (blue) tracks particle $f_\mathrm{v}$.
• Figure 5: Measured time-histories of CH$_4$ (a, e), C$_2$H$_4$ (b, f), and C$_2$H$_2$ (c, g) compared to kinetic model predictions at $T_5$ = 1984 K (a-d) and 2247 K (e-h). The shaded region around each measurement indicates experimental uncertainty. Kinetic models FFCM-2, CRECK, and CALTECH show good agreement with the primary gas products measured. Model predictions of PAH concentrations involved in particle inception and surface growth are subject to order-of-magnitude variations as demonstrated by naphthalene (C$_{10}$H$_8$) in d, h.