Nanofluidic replication of reservoir pore networks on silicon chips: New insights of the miscibility of CO2–alkane systems in nano-confined spaces

The supercritical properties of CO 2 and its carbon neutrality properties drive its wide application in geological utilization and storage, but the complex structural characteristics of reservoir nanopores restrict the utilization of CO 2 . In this study, we innovatively integrate molten wood’s metal perfusion technology and the formation nanopore reproduction process, constructing a nanosilicon-based chip (characteristic scales: 1μm, 200 nm, 100 nm, 30 nm) based on the topological characteristics of the pore network of shale reservoirs.We systematically reveal the dynamic miscibility behavior of CO 2 and seven alkane components (C 6 -C 12 ) within confined space. The experimental results indicate that the nanoconfinement effect significantly reduces the minimum miscible pressure (MMP) of the system, with the maximum reduction reaching 4.18 % at 30 nm (0.68 % and 2.56 % at 200 nm and 100 nm, respectively). The miscible component intervals are also scale-dependent and can be classified into three characteristic regions based on MMP: below C 6 (fast miscible zone), C 7 -C 11 (transitional miscible zone), and above C 12 (hysteresis miscible zone). Furthermore, for the first time, the component differentiation phenomenon of mixed alkane systems was understood in homogeneous and heterogeneous fracture-pore systems. We propose a coupled approach of "digital replication of reservoir pores and in-situ nanofluidic observation" to reproduce the miscible process between CO 2 and key components of crude oil. This novel method, along with robust experimental data, is expected to serve as a valuable reference for the further development of CO 2 geological utilization. • A novel method using Wood's metal injection was applied to characterize shale and create models of single tubes and fracture networks. • Key components affecting the miscibility of CO 2 and shale oil at the micro-nano scale were identified. • Flooding mechanisms in heterogeneous and homogeneous network models matching the shale were explored.