: Carbon Membranes for Natural Gas Sweetening
Evangelos P. Fawasa, Sotirios P. Kaldisb and Xuezhong Hec
"Membranes & Materials for Environmental Separations Laboratory, Institute of Nanoscience and Nanotechnology, NCSR
bChemical Process & Energy Resources Institute, Centre for Research & Technology Hellas 1Department of Chemical Engineering/Faculty of Natural Sciences, Norwegian University of Science and Technology (NTNU)
The first studies of separation membranes began in the eighteenth century (1748) when the Abbe Jean-Antoine Nollet, a French experimental physicist, proved that the bubbling phenomenon in decompressed liquids was caused by dissolved air. In 1824, Rene-Joachim-Henri Dutrochet, a French physiologist, discovered the phenomenon of osmosis in natural membranes . In 1855, Adolf Eugen Fick, a German physiologist, introduced Fick's law of diffusion, which describes the diffusion of a gas across a fluid membrane . In 1861, Thomas Graham, a Scottish chemist, studied the diffusion of gases and laid the foundation of gas and vapor separation through polymeric membranes . He could be called the father of modern dialysis, which is the usage that occupies the biggest share of the membrane market. In 1887, Jacobus Henricus van't Hoff, a Dutch physical and organic chemist, proposed the famous van't Hoff equation for osmotic pressure (n). This work was awarded the first Nobel Prize in Chemistry, in 1901 . After these fundamental works, many other pioneering papers were published during the first half of the twentieth century; the "golden age" of membrane technology (1960-1980) began in 1960 with the invention by Loeb and Sourirajan  of the first asymmetric integrally skinned cellulose acetate reverse osmosis membrane.
Nowadays, together with water treatment/cleaning processes and hemodialysis processes, gas separation processes using membranes are a top priority of academic and industrial interest because natural gas (NG), which exists in deep underground reservoirs, usually contains several non-hydrocarbon components that must be separated and removed. Two of these are hydrogen sulfide (H2S) and carbon dioxide (C02). This process is called "sweetening"  and is one of the major separations in the NG industry The huge amounts of acid-pumped NG demand efficient solutions that are cheaper than existing methods. For example, the emirate of Abu Dhabi alone is expected to produce up to 6.8 billion cf/d of NG by 2020 ; the crude NG in this area contains more than 15% H2S and sometimes up to 50% or even higher . This is an unfortunate fact that many other producer countries also have to confront, which is why many research projects have been funded worldwide in an attempt to solve this major problem.
H2S removal processes can be either physical-chemical or biological. If, for instance, the aim is to produce liquified NG or to remove N2 cryogenic processing, then C02 must be reduced to less than 50 ppm in order to avoid solidification in exchangers, pipes or turbo expanders . H2S, in the presence of water, forms a weak, corrosive acid that causes premature failure of valves, pipelines and pressure vessels. CO, is also corrosive in the presence of water and lowers the heating value. When NG is used as domestic fuel, it becomes necessary to remove H2S because of the health hazards associated with it. The threshold limit value for prolonged exposure of H2S is 10 ppm .
Until now, three main technologies for NG sweetening have been applied
The H,S-rich concentrated acid gas is routed to a sulfur recovery unit to be converted into elemental sulfur by the well-known Claus process, which was first brought on-stream in 1973:
2. Adsorption: Acid gases and water can be effectively removed by physical adsorption on synthetic zeolites. Applications are limited because water displaces acid gases on the adsorption bed .
In these processes, the separation step entails the formation of molecular complexes that must be reversed through a significant increase in temperature. The heating and subsequent cooling of sorbents to prepare them for the next sorption cycle is thus highly energy consuming. In addition, diffusion limitations result in slow uptake and regeneration kinetics, rendering these systems inefficient to process gas with large amounts of H,S and C02.
3. Membranes: Membranes can be used to remove bulk C02 and H2S, preferably at high feed pressures. Different gases pass through the membrane and, depending on their different permeabilities (P), separation is achieved. The main mechanism of gas separation is therefore the difference in diffusion rate of each gas through the membrane, which depends on the sorption and/or molecular sieving effect . In comparison with the other NG separation techniques, the membrane process offers a viable energy-saving alternative since it does not require any phase transformation.