Beyond Conventional Pyrolysis: Addressing Matrix Effects in Recent Marine Sediments

Recent marine sediments contain thermally labile organic matter (OM) embedded in complex mineral matrices that can strongly affect open-system pyrolysis responses, challenging conventional interpretations of source-rock parameters. Here, we evaluate how analytical protocol and sediment matrix influence pyrolysis signals in late Quaternary sediments from the Amazon Fan. Twenty-one samples from seven piston cores collected during the Amaryllis-Amagas Oceanographic Mission were analyzed using three whole-rock pyrolysis approaches: classical pyrolysis (widely known as Rock-Eval), a low-temperature protocol adapted for recent sediments, and the HAWK Petroleum Assessment Method (HAWK-PAM). The same samples were also analyzed after kerogen concentration to provide a matrix-reduced reference. Pyrolysis results were integrated with major- and trace-element geochemistry using Spearman correlations, principal component analysis (PCA), and partial least squares (PLS) regression. Whole-rock thermograms differed systematically among protocols. Classical pyrolysis produced comparatively high S1 responses and poorly resolved S2 peaks, including recurrent shoulders that affected Tmax determination. Lowering the initial pyrolysis temperature substantially reduced S1 but did not fully resolve S2-related interference. In contrast, HAWK-PAM separated low-temperature volatile fractions from the higher-temperature Oil-4 and K-1 fractions, producing the clearest distinction between labile compounds and kerogen cracking. Among whole-rock approaches, HAWK-PAM most closely reproduced trends observed in concentrated kerogen. Kerogen concentration also substantially modified HI and OI relationships, confirming that conventional whole-rock Van Krevelen classification can be strongly matrix-dependent in these recent sediments. Multivariate analyses further demonstrated that pyrolysis parameters respond to distinct inorganic controls. S1 and low-temperature fractions primarily reflect provenance, grain-size/mineral packing and sorptive effects, whereas S2/K-1 covaries with redox-sensitive proxies and transition metals. Tmax is also associated with terrigenous and oxide-related gradients, indicating that apparent maturity may partly reflect matrix effects rather than OM thermal evolution alone. These results demonstrate that protocol choice is critical for pyrolysis of recent marine sediments. HAWK-PAM provides improved separation of thermally labile fractions and kerogen cracking, while inorganic geochemistry offers an independent means of identifying matrix controls. Integrating adapted pyrolysis protocols with mineral-matrix characterization therefore improves OM assessment in young, mud-rich sedimentary systems.