Membrane stripping: desorption of carbon dioxide from alkali solvents

© 2010 Dr. Michael SimioniSolvent absorption of carbon dioxide and other acid gases is a technology that has dominated the gas treatment industry. Solvents offer good selectivity for acid gases as well as allowing for sufficient throughput to capture large volumes of carbon dioxide. Carbon dioxide capture systems are increasingly relevant to a carbon constrained world. Their applicability for reducing emissions from stationary sources such as coal fired power stations is driving research interest. Packed columns currently dominate as the contactor of choice for solvent stripping. However, limiting factors for the implementation of this technology is its size and large capital cost. Packed columns also suffer from many problems systemic of their design and operation. Conditions that deviate from their design often result in reduced mass transfer performance, commonly associated with flooding, foaming and entrainment. With reference to mass transfer, problems may arise if there is not consistent wetting of all the packing within the system. Channelling is an obvious cause for such mass transfer reductions. Membrane contactors offer themselves as an alternative to more conventional contactor systems. The membrane provides a semi-permeable barrier between the liquid and gas streams. This takes the place of traditional packing and prevents direct contact of the two fluids seen in packed columns. Membrane contactors are beneficial in that they offer packing efficiencies up to 30 times greater than that of column systems. Membrane contactors can be made of a wide variety of materials: inorganic (eg: ceramics, sintered metals and glass) and organics (eg: polypropylene (PP), polytetrafluoroethylene (PTFE), polysulfone (PSf) and polyethersulfone (PES)). The main selection criteria for a membrane material for high temperature solvent desorption are hydrophobicity, pore size, chemical and thermal resistance, ease of manufacture, cost and a high solvent breakthrough pressure. The focus of this research is to investigate and develop membranes to meet these selection criteria for use in potassium carbonate solvent stripping applications. Plasma sputtering of PTFE onto microporous hydrophilic nylon and polysulfone membranes has been investigated as one approach. Fluorination of surfaces is known to enhance hydrophobicity by reducing dipole moments. The sputtered membranes produced here show comparable surface hydrophobicity and breakthrough pressures to membranes made purely from more expensive PTFE (static contact angle 120±5°, surface fluorine concentration 70±6%). Plasma treated nylon and polysulfone achieved high static contact angles (nylon/150±5° & PSf/125±5°) and surface fluorine concentrations (nylon/56±6% & PSf/60±6%) optimised at treatment times of 30 minutes and power ratings of 100 W. However, to gain optimal breakthrough pressure (nylon/100 kPa & PSf/85 kPa) extended treatment for 90 minutes at 200 W was required in order to compare favourably with PTFE (88 kPa). This effect is more noticeable in nylon membranes where membranes instantly wetted for pressures as little as 3 kPa at less severe sputtering conditions. These findings suggested that resistance to solvent wetting is not only described by the membrane surface properties, but requires consideration of internal pore surfaces. (Open document for complete abstract)