Sepsis is a blood bacterial infection that can lead to death in few hours in the most severe cases. In sepsis patients, bacteria are difficult to identify because of their low blood concentration (1CFU/ml). Today, sepsis diagnosis requires blood culture for several days before mass spectrometry (MS) analysis of bacteria. Our aim is to develop chemically functionalized porous silicon surfaces (pSi) to enhance the selective trapping of bacteria from blood and allow their identification by infrared desorption-ionization on silicon (IR-DIOS)-MS in few minutes at the bedside of the patient. Compared to UV lasers classically used in DIOS-MS (detection of metabolites), IR will broaden the range of masses detected, which should be more suitable for the analysis of bacteria. pSi surfaces (pore diameters around 1μm) are expected to act as a blood filter, to increase the specific surface, and to improve bacteria desorption thanks to nanophotonic effects. Bacteria will be specifically trapped on the pSi surface thanks to the immobilization through copper free click reaction of azide modified nanobodies with dibenzocyclooctyne (DBCO) modified surfaces. Our first objective is to functionalize the pSi surface with DBCO-PEG-silane molecules in vapour-phase [1]. This method enables the high aspect ratio surfaces to be functionalised homogeneously, without any solvent, and to mix silane molecules with different chain lengths and with/without DBCO terminal groups. A home-made reactor was built and the impact of process parameters on propyldimethylmethoxysilane was first investigated (pSi surface temperature, silane molecule injection modes, silanization duration). The functionalized surfaces were characterized by FTIR-ATR. The process is being implemented for DBCO- PEG34-silanes molecules. The availability of DBCO groups, as well as the conservation of their structure, are essential to allow the subsequent grafting of a nanobody. In addition, the nanobody must maintain its bioactivity with respect to antigens. Therefore, its structure must not be modified through interaction with the surface. Consequently, molecular dynamics (MD) simulations were performed to predict the optimal (i) surface coverage of DBCO-PEG-silane molecules, (ii) PEG chain length, (iii) the “DBCO-PEG-silane: PEG-silane” ratio, and (iv) the PEG chain length of the DBCO-PEG-silane molecules in a mix self-assembled monolayer. The availability of DBCO was found to be optimal under the following conditions: 0.3 molecule per nm² and a ratio “DBCO-PEG34-silane: PEG34-silane” equal to 1:1 (Fig.1). We are now studying by MD simulations the conformational changes of the grafted nanobody under such conditions. Future work will focus on establishing the proof of concept of our device by grafting Listeria nanobodies on the DBCO-PEG-silane modified pSi surfaces to trap Internalin B in model solution and then in plasma for IR-DIOS-MS analysis.