Graphene-based membranes – a new opportunity for CO₂ separation

Keywords:

• graphene
• graphene oxide
• membrane
• gas separation
• CO₂ capture
• post-combustion

Currently, membrane separation processes play a very significant role in the separation industry. Widely used membrane processes include microfiltration, ultrafiltration, nanofiltration, reverse osmosis, electrolysis, dialysis and gas separation. In particular, membrane-based gas separation is commonly used for O₂/N₂, H₂/N₂, H₂/CH₄, H₂ recovery from ammonia plants or in oil refinery processes, CH₄ separation from biogas, CO₂/CH₄, and volatile organic compound (VOC) removal. Among these, CO₂/N₂ separation has been a hot issue in membrane fields because of an urgent need for CO₂ capture from fossil-fuelled power plants, in industrial processes and in the fuel production and transformation sectors, to prevent global warming and climate change. But what are challenges for CO₂ capture? CO₂ capture can be done with technologies presently available, but this will increase power generation costs, will require more fossil fuel for the same power generation capacity and will increase our reliance on fossil fuels, increasing the supply security concerns. On top of this, there is no experience with CO₂ capture at the power plant scale.

CO₂ capture technologies can be classified into either process or technology platform/components. The process component includes post-combustion decarbonization (conventional power plants), pre-combustion decarbonization (new power plants) and denitrogenation (new power plants). The technology component includes membranes, absorption, adsorption, cryogenics, carbon extraction, biotechnology and energy conversion. Undoubtedly, the best technology for individual CCS applications depends on the power plant and its fuel characteristics. A current leading technology for post-combustion CO₂ capture is amine absorption. Indeed, amine absorption is a proven, mature technology for separating CO₂. However, the cost of CO₂ capture from flue gas with amines is still high (US$40–80/t CO₂), while the US Department of Energy target requires below $30/t CO₂ [1]. In addition, amine energy requirements are also high (>30% of the power plant production), while the US Department of Energy target is <20% parasitic energy loss. Amines are adversely affected by oxygen and SO₂, and amine adds 3 m³/h/MW in water usage [2]. Amine plants cover ∼5 acres; therefore, large numbers of towers are needed so that the scaling advantage of absorption processes disappears.

In this regard, membrane-based gas separations offer great potential as an energy-efficient, low-cost CO₂ capture option [3]. However, to compete with conventional technologies, membrane materials and modules to treat large volume of flue gas should be developed. In general, the high cost of separating CO₂ from power plant flue gas is needed because of the low partial pressure of CO₂ (∼0.13 bar) and the enormous volume of gas emitted from a typical power plant. For example, an average 600 MW_e coal-fired power plant emits 500 m³/s (1540 MMscfd) of flue gas containing ∼10–13% CO₂, which amounts to ∼11,000 t CO₂/day [3]. This means high CO₂ permeances (or fluxes) should be needed in terms of both membrane and process properties to reduce the size and number of membrane module. According to a recent report on CO₂ capture cost by membrane processes, if membrane materials with >4000 gpu (gas permeation unit, 1 gpu = 1 × 10⁻⁶ cm³ (STP) cm/cm² sec cm Hg) (CO₂ permeance) and >30 (CO₂/N₂ selectivity) can be developed, the capture cost approaches to $15/t CO₂ [3]. Although there have been lots of advances in membrane-based gas separation, there is yet to be membrane materials that meet these high criteria.

Graphene is the world's thinnest, two-dimensional material – a single-atomic-layer honeycomb lattice of carbon atoms in an sp² hexagonal bonding configuration. Owing to its two-dimensional extension in the macroscopic range and atomic thickness, graphene is highly being considered as an ideal membrane platform, if pore size and porosity can be properly engineered since graphene sheet is actually impermeable [4]. A number of theoretical and experimental studies have been performed to develop porous graphene membranes [5–7]. According to more recent studies, graphene-based membranes may soon come into reality [8,9]. Graphene was first isolated in 2004. Although there has been a boom of graphene research lately, no efforts have been put into analyzing its usefulness as a gas separation membrane. With the advent of graphene age, large-area, monolayer graphene sheet can now be produced by using chemical vapour deposition as well as graphene or graphene oxide (GO) nanosheets from chemical exfoliation of graphite. That is, graphene as a raw material source is easily available.

For gas separation through molecular sieving (i.e., size exclusion), subnanometric pore sizes should be created on graphene basal plane (note that the kinetic diameters of CO₂ and N₂ are 0.33 and 0.364 nm, respectively), but such technologies are not yet available and will be a great challenge in the future. Currently, irradiation of strong electron beam is used to make holes on the graphene membrane, but the pore sizes is still too large to separate gas molecules selectively. Alternatively, some trials to prepare membranes by using GO or reduced GO nanosheets have been recently reported [8]. Intrinsically, GO is a highly oxidized graphene sheet, produced by strong oxidation of graphite. It is hydrophilic and has many oxygen-containing polar groups (e.g., –OH, –C–O–C– and –COOH) on the basal plane and at the edges. As such, GO can be easily dispersed in aqueous solution. The key idea to use GO nanosheets as a basic building block for membrane construction is to create gas-permeable channel between two-dimensional GO sheets. Interplanar spacing of GO is usually 0.6–1.0 nm [8], depending on the preparation conditions and the presence of intercalated water molecules, while that of graphite is 0.335 nm, that is, no gas diffusion occurs between interspacing. More recently, few-layered GO membranes have been shown for CO₂ separation. These GO membranes were coated onto commercially available microporous polymeric membranes by spin-casting or spray-coating methods. Due to unique GO structure, GO stacking structures can be tuned by different coating methods, and the thicknesses are as thin as ∼3–5 nm. Also, GO nanosheets show a high CO₂ sorption capability due to its polar groups and hydrophilic nature, leading to selective CO₂ separation properties in both dry and hydrated states. For instance, at a hydrated state (relative humidity of 85%), the CO₂/N₂ selectivity increases to 60 with increasing CO₂ permeance. Since many industrial gas streams such as post-combustion, natural gas sweetening and syngas contain water vapour, the effect of water vapour on membrane separation performance is very crucial. Commonly, water vapour strongly deteriorates membrane performance; with reduction of both permeability and selectivity by water condensation on membrane surfaces or pores, energy-costly water vapour removal before membrane unit is necessary. In this regard, water-enhanced CO₂ separation in GO membranes will be of course a great benefit in post-combustion capture design because dehydration by condensation in the permeate side is easier than dehydration at high-pressure feed side.

In summary, CO₂ capture from power plants is essential, but no technology is a clear winner. Membranes have some advantages over conventional technologies, and future processes for the generation of power and fuels, combined with the need for CO₂ sequestration, will create significant opportunities for membrane-based gas separations, including CO₂ capture at conventional power plants and advanced power plants (integrated gasification combined cycle), hydrogen production by reformer processes and CO₂ capture in biofuel production processes. In order to apply GO or other graphene-based membranes to practical operations, the following characteristics need to be achieved: more high CO₂ permeance (high-flux support membranes should be developed to achieve it) and selectivity under practical operation conditions (mixed gases in the presence of water), life time of at least 1 year, tolerance to other impurities (e.g., SO₂, NO_x and trace metals), temperature, pressure, etc. These requirements remain to be fulfilled. The future work with this system will be directed at clarifying more exact transport and separation mechanism, characterization of the GO or graphene-based membranes, development of roll-to-roll fabrication processes and achieving the optimum operation conditions.

Additional information

Notes on contributors

Ho Bum Park