Sepsis, a blood bacteraemia infection, is one of the first mortality cause in hospitalized patients. Sepsis diagnosis are currently based on the identification of the bacteria by MALDI-MS analysis. However such analysis require long experimental preparation protocol including blood culture during several days. As survival chances decline by 7% each hour, wide range antibiotics are administrated, in first instance, increasing antibiotics resistance emergence probability. So rapid and sensitive sepsis diagnostic tools are urgently required. Recently, a pattern of blood metabolites (including amino acids, phospholipids and acids) have been identified as sepsis biomarkers. However metabolites are low molecular weight species present in trace in blood among an abundant background of high molecular weight species. So the detection of specific metabolites for clinical use is still a challenge. In a previous study we have shown that nanoporous silicon (pSi) surfaces allow to trap metabolites from a plasma sample, while being compatible with MALDI-MS detection. Nevertheless the detection of metabolites was not sensitive enough for sensitive diagnosis of pathological samples, likely because of a low affinity between metabolites and the silicon surface. In the present study, we have investigated the efficiency of surface chemical functionalisation of pSi, by (with?) silane molecules, for sepsis metabolites harvesting from plasma and and the detection by MALDI-MS of the trapped metabolites detection. Four different silane molecules were studied, with different alkyl chain lengths and head-group charges, as well as 20 sepsis and common blood metabolites. Firstly silanization protocols of pSi surfaces, in gas phase, were elaborated. ToF-SIMS, reflectometry, and XPS analysis suggest that the silane monolayers are conform into the pores (10nm x 1µm). Then a MALDI-MS protocol was developed to analyseis metabolites trapped into the functionalized pSi substrates from model solutions. Results suggest that silane monolayers allow to significantly improve the trapping of several metabolites (in comparison with bare pSi) and that specific metabolites can be harvested by tuning the nature of the silane monolayers. Also, we show that silane monolayers lead to a significant increase in the DIOS-MS signal of harvested metabolites. So, silane monolayers are promising candidates to improve the sensitivity of sepsis metabolites detection. This methodology is currently and are being evaluated in using serum samples. Moreover Molecular dynamics simulations were developed to investigate at the atomic scale, for the first time, the interactions between metabolites and silane monolayers. The interactions energies between silane layers and metabolites are in line with the MALDI-MS results. So our MD simulation protocol could help to optimize and predict metabolites trapping, but also to understand more fundamental aspects like the effect of the organization of interfacial water molecules on the adsorption of metabolites.