3DLung - 3D-printed Membranes for Artificial Lungs


Felix Hesselmann


+49 241 80 47059




State of the art membrane lungs consist of thousands of hollow fiber membranes, which separate the blood from the gas phase and offer a high surface area required to overtake the functionality of the human lung in case of an acute or chronic respiratory disease. Typically, blood flows around the fibers while gas is lead through the inner cavity of the hollow fibers. The blood is oxygenated and decarbonized through diffusion of each gas over the membrane wall.

Although, membrane oxygenators are routinely used in cardiopulmonary interventions and due to a lack of alternatives at a progressive rate for the therapy of chronic lung disease, three factors mainly prohibit the long-term use or the often envisioned implantation of artificial lungs: 1) the limited gas transfer caused by the blood sided resistance, 2) the thrombogenictiy of the devices due to inhomogenous flow distribution and 3) design restrictions given through the non-organic design of hollow fiber membranes.

New 3D-printing technologies allow all degree of freedom for the manufacturing of novel membrane architectures and, thus, promise to overcome the stated limitations.


The objective of this research project is to evaluate the advantages of 3D-Membranes for artificial lungs. Therefore, we will proof the and further evaluate the potential of 3D-membranes to overcome the mentioned limitations of as a significant step towards an implantable artificial lung.

In order to meet the objective and guarantee a the evaluation of the 3D-mmebrane lung up to a clinical level, this research project benefits from the interdisciplinary collaboration of the Department of Cardiovasular Engineering (CVE) with the Leibniz Institute for interactive materials (DWI) and the Department of Internal Medicine (Section for Pneumology).


The responsibilities of the CVE within this research project lays in the design work for a 3D-membrane module including a housing and the subsequent in vitro testing.

The three limitations stated above will be addressed directly within the design evaluation phase. Gas transfer in artificial lungs is limited through the blood-sided resistance. Passive laminar vortex mixing is one idea to decrease the blood-sided diffusive resistance. By overcoming this resistance, a diffusion coefficient closer to the membrane limit seems achievable. A more efficient gas exchange promises less required exchange area, and consequently high potential for miniaturization and a lowered risk of thrombus formation.

For this purpose, multi-objective numerical simulation of flow and gas transfer using computational fluid dynamics (CFD) are conducted. This offers the possibility to find an optimal geometry to enhance gas transfer but not cause any additional blood trauma at the same time.

The already mentioned risk of thrombus formation as the main limiting factor for long-term applications is partly of a flow-induced nature in many cases and caused by stagnation zones paired with zones of elevated shear rates. While a hollow fiber module has one constant blood flow permeability, 3D-membranes allow adapting the permeability locally, e.g. by reducing the flow resistance on a microscopic level. Hence, it is possible to allow blood flow in the areas where typically stagnation occurs and vice versa reduce the flow velocity in zones with high shear rates. Therefore, washout-simulations using CFD help to identify these areas of interest and iteratively adapt the permeability until a homogeneous flow distribution is reached.

The subsequent in vitro testing requires a design of an up-scaled laboratory sample for the evaluation of the gas transfer performance and hemolysis.

The manufacturing of solid membrane made of silicone was already proven feasible prior to this research project on a small scale. The DWI investigates the processing of porous and further integral asymmetric high-flux membranes with a dense outer skin, which represent stand-of-the-art for hollow fiber membranes utilized for prolonged applications.

Cooperation Partners
Funding This project is funded by the German research foundation, project number 347368182