Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 production under pre-combustion operating conditions

G Model CATTOD-8717; No. of Pages 9

ARTICLE IN PRESS Catalysis Today xxx (2013) xxx–xxx

Contents lists available at ScienceDirect

Catalysis Today journal homepage: www.elsevier.com/locate/cattod

Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 production under pre-combustion operating conditions ˜ ∗ , M.M. Barreiro, Y. Torreiro, J.M. Sánchez M. Marono Centre for Energy, Environmental and Technological Research (CIEMAT), Department of Energy, Av. Complutense, 40, 28040 Madrid, Spain

a r t i c l e

i n f o

Article history: Received 31 July 2013 Received in revised form 14 October 2013 Accepted 1 November 2013 Available online xxx Keywords: Hybrid systems Sorption enhanced WGS reaction CO2 capture WGS membrane reactor H2 production

a b s t r a c t This paper presents experimental results obtained for different approaches used for the simultaneous capture of CO2 and production of H2 under precombustion conditions. A novel hybrid system adsorbent–catalyst–membrane has been designed at Ciemat which combines a sorption enhanced WGS process (which consists of a mixture of adsorbent and catalyst) with a hydrogen selective membrane. First, the individual performance of the membrane and the binary adsorbent–catalyst system is evaluated in terms of H2 permeability, carbon monoxide conversion and CO2 capture and best operating conditions are determined for both approaches. The crucial role of temperature and steam on the overall performance of the system is especially discussed. Finally, first experimental results obtained for the hybrid adsorbent–catalyst–membrane system are presented and the advantages and challenges of adding a H2 selective membrane to the sorption-enhanced process are analysed. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Some of the most promising alternatives to the use of liquid amines and PSA for the capture of CO2 and H2 production in precombustion processes are the use of solid sorbents for capturing CO2 and the use of selective membranes for the separation of H2 . Different approaches and process schemes, based on those upand-coming technologies or on a combination of them, are being investigated by several authors to improve the capture of CO2 and to increase the production of H2 in gasification processes. In the field of sorbents and catalysts development it is worth mentioning, among others, the efforts devoted to the preparation of new adsorbents with improved CO2 capture capacity and tolerance to sulphur at intermediate temperatures for their application to WGS processes, the preparation of hybrid materials based on CaO for in situ capture of CO2 at high temperatures or the development of bifunctional adsorbents capable to simultaneously catalyse the WGS reaction and capture CO2 [1–3]. Regarding membranes technology for H2 production and/or separation of H2 /CO2 mixtures great advances are being gathered in the development of very thin Pd or Pd-alloy based membranes and in the design of new membrane supports based on ceramic or metallic hollow fibres, nanoporous hybrid silica membranes or

∗ Corresponding author. Tel.: +34 913466021. ˜ E-mail address: [email protected] (M. Marono).

novel non Pd metallic membranes such as vanadium membranes. A detailed review of the most recent developments in membranes and membrane reactors for CO2 capture and H2 production can be found in the literature [4,5]. Different hybrid system, based on the combination of several of the above mentioned technologies are also being studied in order to take advantages of the possible synergies between them. Two of these approaches are the WGS membrane reactor approach (WGSMR) or the sorption enhanced Water Gas Shift (SEWGS) process. In a WGS membrane reactor a WGS catalyst is combined with a H2 selective membrane. The continuous removal of H2 from the reaction media by permeation through the membrane enhances the conversion of CO to CO2 and H2 . This approach requires the use of a H2 highly selective membrane and up to now dense Pd and Pd alloy based membranes can be considered the most advanced ones. This approach is being investigated by several research groups [6–8]. The combination of a CO2 sorbent and a WGS catalyst in a WGS unit is known as the sorption enhanced WGS process (SEWGS) and it allows an increase in the conversion of CO due to the removal of one of the products (CO2 ) by the sorbent. The WGS process is industrially performed using two adiabatic converters, one of them at high temperature using an iron–chromium based catalyst followed by a second WGS reactor usually run at significantly lower temperature. The SEWGS approach provides the opportunity to reduce the energy penalty associated to the cooling process between reactors due to the use of a single reactor that operates at

0920-5861/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cattod.2013.11.003

˜ et al., Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 Please cite this article in press as: M. Marono, production under pre-combustion operating conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.11.003

G Model CATTOD-8717; No. of Pages 9

ARTICLE IN PRESS no et al. / Catalysis Today xxx (2013) xxx–xxx M. Maro˜

2

high temperature. Suitable high temperature WGS catalysts to be used in a SEWGS process should be highly active at temperatures in the range of 300–500 ◦ C, show resistance to sintering at temperatures above 450–500 ◦ C and maintain a stable performance under cyclic operating mode. For their application to SEWGS processes, suitable sorbents are required to have high CO2 capture capacity and selectivity towards CO2 at temperatures in the range of 300–500 ◦ C, adequate sorption–desorption kinetics (easy regenerable), mechanical strength and low cost [9–11]. Only a few types of materials are reported in the literature to be able to meet the above mentioned objectives and among them hydrotalcites have demonstrated to be suitable to be used in processes such us the SEWGS process [12,13] or the sorption enhanced steam methane reforming [14,15]. Recently the possible effects of the presence of contaminants such as H2 S are also being investigated [16]. However, both approaches (WGSMR and SEWGS) allow only for removal of one of the products of the WGS reaction (H2 or CO2 , respectively) and usually the resulting reaction rate enhancement is not sufficient to achieve significant conversion at low temperatures. The opportunity to take advantage of the synergies that can be created by the combined removal of CO2 and H2 has motivated the development of a novel hybrid system sorbent–catalyst–membrane that has been designed at CIEMAT. The whole system is known as the sorption enhanced water gas shift membrane reactor (SEWGSMR). To the best of our knowledge, very few references can be found in the literature on the performance of hybrid adsorbent–catalyst–membrane systems for WGS applications as for example the WGS-HARM hybrid system proposed by Harale et al. [17] so research in this direction is still needed. The novel hybrid system developed at Ciemat consists of a high CO2 capture capacity sorbent, a high temperature commercial WGS catalyst and a dense Pd membrane highly selective to H2 . The selection of the different elements (sorbent, catalyst and membrane) that shaped the proposed hybrid system was done based on the following criteria: compatibility of best operating conditions, degree of development of the technology and availability and they were selected based on our previous works [18–21]. Using those criteria a K-doped hydrotalcite with high CO2 capture capacity under WGS operating conditions, a commercial Fe–Cr high temperature WGS catalyst and a dense Pd-based membrane highly selective to H2 .were selected as the most suitable to build the hybrid system. A bench-scale Pd membrane reactor, available at Ciemat, was used as the base case for the design of the new hybrid system. In this paper the influence of main operating parameters (pressure, temperature, volume ratio adsorbent/catalyst and steam) on the performance of the proposed hybrid system is evaluated and cyclic behaviour was tested using four sorption–desorption cycles. Main advantages and limitations of the proposed novel system versus the other two approaches (SEWGS and WGSMR) are also discussed.

rate is 4.5 nL/min. The unit can work at up to 700 ◦ C and 30 bar. Desired gas mixture is produced synthetically using mass flow controllers (Hi-Tech) and deionised water is metered by a piston pump (Gibson 307) and vaporised before entering the reactor. A more detailed description of this reaction system can be found elsewhere [19]. The H2 separation experiments and the hybrid system adsorbent–catalyst–membrane tests have been performed at bench-scale in a Pd-based membrane reactor facility in which a gas flow-rate of up to 2 Nm3 /h can be treated at a maximum temperature of 1200 ◦ C and up to 20 bar. A schematic of this experimental setup is shown in Fig. 1. The facility consists of a tube and shell membrane reactor that holds a tubular Pd-based membrane supplied by CRI Catalyst Co., USA. It has local control and on-line remote control, based on R485 digital communications. The process is automatically monitored and the main operating parameters are continuously recorded. Pure gas components (H2 , CO, CO2 , and N2 ) are supplied from gas cylinders to the mixture preparation section. Steam is produced by feeding water to an electrically heated coiled tube vaporiser by means of a displacement pump (Dosapro). Dry gas and steam are mixed before entering the reactor. Permeate and retentate gas flow rates are measured with mass flow metres. An additional sweep gas line is prepared to use N2 or steam as sweeping gas if necessary. A more detailed description of the facility can be found elsewhere [21]. For both series of studies, inlet gas, exit gas, permeate and retentate gas compositions are measured by gas chromatography using a CP4900 Varian gas chromatograph equipped with two columns, a Porapack and a molecular sieve column and with two thermal conductivity detectors. 2.2. Sorbent A potassium carbonate doped hydrotalcite based material was used in these experiments. This sorbent was selected in our previous work [18] among different types of materials as the most appropriate, in terms of CO2 capture capacity and regenerability, to be used under conditions of WGS processes (temperatures in the range of 300–500 ◦ C and presence of steam). It was supplied by SASOL (Germany) in pellets of 5 mm × 5 mm. The sorbent was prepared by calcination of the supplied material at 600 ◦ C for 4 h. 2.3. Catalyst

2. Experimental

The catalyst used in this work is a commercial Fe–Cr high temperature water gas shift catalyst that proved to be highly active at temperatures above 350 ◦ C [20]. This catalyst is currently being used in the 14 MWth pre-combustion CO2 capture plant in Puertollano IGCC plant [22] and according to the information provided by the manufacturer, Johnson Matthey, chemical composition of the catalyst is Fe2 O3 88 wt%, Cr2 O3 9 wt%, and CuO 2.6 wt%. [23]. For using it in the proposed hybrid systems, the performance of the catalyst in the temperature range of 250–400 ◦ C was studied in detail in this work. The catalyst was supplied in pellets of 6 mm × 3 mm and the amount used in each test was 5 g.

2.1. Test rigs

2.4. Membrane description

Two different experimental test rigs have been used in the work presented in this paper. The experiments carried out to test the sorbent, the catalyst and the binary system adsorbent/catalyst were performed in a Microactivity Pro lab-scale Unit. It is an automatic and computerised laboratory rig that consists of a stainless steel tubular reactor of OD 12.8 mm, ID 10 mm and 380 mm long housed in a one single zone SS304 oven. The maximum operating gas flow

As mentioned previously, the membrane used in this work is a pre-commercial prototype supplied by CRI Catalyst Co., USA. The membrane consists of a 0.012 m2 active surface area composite Pd-based membrane prepared on a porous stainless steel tubular support (O.D. = 2.54 cm; L = 15 cm) welded to a 316 SS tube which is closed in one end. According to the supplier the membrane provides a permeance in the range of 20–40 Nm3 /m2 h bar0.5 . The whole

˜ et al., Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 Please cite this article in press as: M. Marono, production under pre-combustion operating conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.11.003

G Model CATTOD-8717; No. of Pages 9

ARTICLE IN PRESS no et al. / Catalysis Today xxx (2013) xxx–xxx M. Maro˜

3

Fig. 1. Schematic of the bench-scale membrane plant used in this work for membrane reactor studies.

assembly is housed in a tubular reactor and usually flowing gas goes from the outside to the inside of the membrane. However, in this work, the configuration of the membrane reactor has been modified to allow an inside–outside permeation flow of H2 in order to avoid deterioration of the palladium layer for direct contact with the solid during the hybrid sorbent–catalyst–membrane studies.

guarantee that complete conversion of CO takes place during the sorption step. Desorption was performed using N2 at 500 ◦ C and 1 bar. A complete cycle of sorption–desorption usually lasted 3–4 h. Both the sorbent and the catalyst were crushed and sieved to a size between 1.6 and 2.3 mm to favour a homogenous mixture in the reactor. 3. Results and discussion

2.5. Sorption–desorption tests procedure All the sorption–desorption tests were performed using a total amount of 10 g of solid (adsorbent alone or a mixture of adsorbent/catalyst) following the same procedure: first the samples were heated in N2 at 500 ◦ C for 1 h before performing the sorption tests in order to remove any species that could have adsorbed on the surface. Then, the reactor was cooled down and pressurised to the desired process temperature and pressure in a N2 flow. The reactor was then bypassed and water, previously vaporised, was fed to the system. Then, N2 flow was changed by the process gas flow and the wet gas mixture was allowed to enter the reactor. Sorption tests were performed at temperatures in the range of 300–400 ◦ C using feed gas mixtures consisting of CO2 /N2 and CO/N2 depending on the experiment. After saturation of the sorbent, the reactor was bypassed again and the facility was heated at 500 ◦ C and depressurized to atmospheric pressure in a N2 flow in order to guarantee the complete desorption of CO2 . When the regeneration conditions were reached, the reactor was connected again and the desorbed species were swept away by a N2 stream and measured by gas chromatography. For the binary system adsorbent/catalyst, different volume ratios (Vads /Vcat ) were used in the tests, from only adsorbent to Vads /Vcat = 10. This parameter must be defined very carefully to

3.1. Catalyst and sorbent performance in the temperature range of interest The performance of the catalyst in the temperature range of interest for its application to SEWGS process has been investigated. Fig. 2 shows that when using the catalyst for the first time (fresh catalyst) the catalytic activity increased with increasing temperatures starting at as low temperatures as 200 ◦ C. However, after heating the catalyst at 500 ◦ C, as required during the regeneration steps of the cyclic SEWGS process, it suffered a drastic reduction in its catalytic activity at temperatures lower than 350 ◦ C. This behaviour showed that, as reported for different high temperature WGS catalyst, partial sintering of the catalyst may take place during treatment at high temperatures which did not affect to the performance in the upper temperature range, where the catalyst is highly active but clearly influences the catalytic activity at lower temperatures where the catalyst is much less active. Several temperature cycles were performed to confirm the stable behaviour of the catalyst under the proposed SEWGS operating conditions. Four cycles of heating at 500 ◦ C and cooling down at 300 ◦ C were performed and the results obtained showed that the catalyst attained a stable catalytic activity versus temperature after the first

˜ et al., Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 Please cite this article in press as: M. Marono, production under pre-combustion operating conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.11.003

G Model CATTOD-8717; No. of Pages 9 4

ARTICLE IN PRESS no et al. / Catalysis Today xxx (2013) xxx–xxx M. Maro˜

Fig. 2. Effect of SEWGS process regeneration temperatures on catalyst performance. Fig. 4. Performance of the sorbent under cyclic operating mode. Feed: 5% CO/N2 , P = 15 bar, T = 350 ◦ C, 50% (v/v) steam.

The kinetics of the WGS reaction favoured that the catalytic activity showed by the sorbent towards the WGS reaction increased with increasing temperature. Fig. 3 shows the effects of temperature in the conversion of CO (XCO ), where XCO was calculated according to Eq. (1) as follows: XCO =

Fig. 3. Conversion of CO versus temperature. Only adsorbent. P = 15 bar; Feed gas composition: 5% CO/N2 ; steam = 50% (v/v).

heating cycle without showing any hysteresis cycle. Fig. 3 depicts the stable catalytic activity of the catalyst versus temperature when tested at 15 bar using a feed gas composition of 5% CO/N2 and 50% steam. It is interesting to note that the catalyst started to show catalytic activity towards the WGS reaction at temperatures higher than 300 ◦ C and only at temperatures above 350 ◦ C the conversion of CO can be considered significant. For that reason, for using the catalyst in the hybrid adsorbents/catalyst system, process temperatures in the range of 300–350 ◦ C were chosen in order to be able to determine the effect of the presence of the sorbent on the conversion of CO (sorption-enhanced reaction effect). The sorbent considered in this work is a potassium carbonate promoted hydrotalcite supplied by SASOL (MG61-K2 CO3 ) which proved, in our previous work [18], to be very suitable as a sorbent for CO2 capture under WGS operating conditions. As mentioned above, the sorbent was prepared by calcination at 600 ◦ C for 4 h which was found to guarantee that all CO2 was released from the structure and the sorbent was ready to capture CO2 [24]. A significant property of this sorbent, found when tested in presence of the main syngas components (CO, H2 , CO2 ) [17] is that the sorbent showed catalytic activity towards the WGS reaction in the temperature range of interest (300–500 ◦ C). The role of the potassium carbonated species in the CO2 capture capacity of the sorbent and the main mechanisms that might be involved is still under investigation [24–26].

[CO]i − [CO]e [CO]i

(1)

where [CO]i is the molar concentration of CO in the feed gas in standard conditions (298 K, 101 kPa), [CO]e is the molar concentration of CO in the exit gas stream in standard conditions (298 K, 101 kPa). This behaviour can be controversial for the SEWGS process because the use of a temperature equal or above 350 ◦ C during the sorption step would improve the conversion of carbon monoxide but in detriment of the capture capacity of the sorbent. Although during the initial contact of the sorbent with the gas stream all CO resulted converted, probably due to the exothermicity of the reaction, the catalytic activity showed by the sorbent towards the WGS reaction was decreasing with decreasing temperature and also with increasing time. Fig. 3 shows that a noticeable increase in the slope of the curves can be observed with decreasing temperature suggesting a significant deactivation of the sorbent with time on stream at temperatures lower than 350 ◦ C. performance of the sorbent under cyclic The sorption–desorption mode was then investigated at laboratory scale in four sorption–desorption cycles at 300 ◦ C and 350 ◦ C. Fig. 4 shows the breakthrough curves obtained at 350 ◦ C. The first very interesting fact that can be observed in Fig. 4 is the demonstration of the so called sorption enhanced reaction: during the sorption step all CO was converted to CO2 and H2 and until the saturation of the sorbent, all CO2 (and also some H2 ) produced was captured by the sorbent. Besides that, simultaneous breakthrough of CO and CO2 took place. Another interesting fact was that the final catalytic activity of the sorbent started to decrease after the first sorption cycle from more than 90% to less than 40%. While the sorbent was capturing CO2 only H2 was present in the exit gas. It is also important to mention that H2 was partially adsorbed during the sorption step but the extent of this process seemed to decrease with cycles. For the cycles performed simultaneous breakthrough of CO and CO2 took place during this test suggesting that despite the decrease observed in the catalytic activity, the remaining activity is still enough to convert all CO until the saturation of the sorbent. Another relevant parameter that must be considered in

˜ et al., Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 Please cite this article in press as: M. Marono, production under pre-combustion operating conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.11.003

ARTICLE IN PRESS

G Model CATTOD-8717; No. of Pages 9

no et al. / Catalysis Today xxx (2013) xxx–xxx M. Maro˜

5

1100 1000 900

700

2

JH2[mol/m .h]

800

600 500

O-I 593 K O-I 653 K O-I 753 K I-O 593 K I-O 653 K I-O 753 K

400 300 200 100 0 0

100

200

300

400

500

600

Ffeed [L/h] Fig. 5. Effect of flux direction on the permeation of hydrogen with temperature and feed flow rate. OI, outside–inside; IO, inside–outside. P = 1.5 MPa.

this process is the amount of steam used. Under SEWGS conditions the presence of steam is required in both processes involved: (1) during WGS reaction to guarantee the selective conversion of CO to CO2 and H2 avoiding the occurrence of non-desired secondary reactions such as methanation or disproportionation reactions [19,27,28], and (2) during the CO2 capture process, to allow the formation of hydrates, carbonates and other intermediate species reported in the literature to be responsible for the CO2 capture capacity of K-doped hydrotalcite based sorbents under conditions of WGS reaction [24,29]. For high temperature WGS Fe–Cr catalysts, as the one used in this work, an optimum H2 O/CO ratio of 2 has been widely accepted while in the case of potassium promoted hydrotalcite sorbents, the use of high values of steam partial pressures have been reported to improve their CO2 capture capacity [24,29–32]. Based on our experience and on the recommendations described above, for the SEWGS experiments carried out in this work a clear excess of steam has been used (50%, v/v) to guarantee a high steam partial pressure during the capture process and to minimise the occurrence of secondary non-desired reaction that could occur if the WGS process is carried out in deficiency of steam. 3.2. Membrane reactor configuration The layout of the existing bench scale membrane reactor was modified to allow the study of the proposed hybrid adsorbent–catalyst–membrane system. For WGS membrane reactor studies usually the catalyst is placed in the annulus of the membrane reactor, in direct contact with the Pd layer of the membrane. In the new configuration the gas is fed by means of a diver in the inner part of the membrane and the solid (adsorbent and catalyst) is placed inside the Pd membrane tube, filling the volume surrounded by the Pd membrane surface. Although this configuration has been used by other authors [33] it has some disadvantages over the former one such as a reduction in the mass of solid that can be used or the negative effect of the driving force acting against the Pd layer. Fig. 5 shows that permeation from the inside to the outside of the membrane reduced slightly the stable maximum flux of hydrogen so this must be taken into account when using this configuration. However, this configuration adds valuable advantages to the hybrid system such as to avoid the surface damage of the Pd layer due to its direct contact with the solid increasing considerably the durability of the membrane.

Fig. 6. Permeation of hydrogen in bynary mixtures: comparison of the effect of N2 , CO2 and steam.

3.3. Membrane permeation studies As described presviously, the membrane used in this study is a dense Pd-based tubular membrane completely selective to H2. According to our previous studies [34] when pure hydrogen was used as feed gas, the permeability of H2 increased with increasing temperatures, following the Arrhenius Law, and increasing driving force, following the Sievert’ Law. However, besides temperature and pressure, the permeation of hydrogen through Pd-based membranes can be affected by many other factors including for example the feed gas flow-rate or the presence of other co-exiting gases in the feed gas stream [35–37]. In this direction, in order to determine the performance of the membrane under conditions of SEWGS process which would be later used in the proposed hybrid system, a series of experiments were carried out to study the effect of the presence of N2 , CO, CO2 and steam on hydrogen permeation and on membrane integrity. Fig. 6 shows some results that summarise the influence of each of these components on the permeation of H2 through the membrane. As it can be observed in Fig. 6, the flux of permeated hydrogen decreased when the concentration of co-existing gases in the feed gas increased and clear differences can be observed between N2 , CO2 and H2 O. For N2 and CO2 the inhibitive effect was proportional to the amount of compound added to the feed gas. This behaviour can be associated to dilution effects. However, in the case of steam, inhibitive effect is less pronounced but it appears at relative low concentrations and remains almost constant or slightly increasing when much higher amounts of steam are used. This behaviour suggested that even at low concentration, steam might be adsorbing on the membrane layer causing a decrease in the flux of H2 which permeated. Moreover, according to the literature, the presence of CO can lead to severe deactivation of palladium-based membranes. One of the main causes that have been reported to occur is the strong adsorption of CO on the Pd surface, which contributes to the deactivation and blocking of hydrogen dissociation sites [38]. This negative influence of the carbon monoxide on the permeation of H2 has been reported to depend on the temperature and on its concentration in the feed gas [39]. As carbon monoxide is necessarily present in the syngas that will be fed to the proposed hybrid system, we tested the influence of the concentration of CO on the permeation of H2 through the membrane. Fig. 7 shows the flux of hydrogen versus the CO contents in the feed gas for different mixtures N2 /CO/H2 at 450 ◦ C. As it can be observed in Fig. 7 the presence of CO in the feed gas has great

˜ et al., Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 Please cite this article in press as: M. Marono, production under pre-combustion operating conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.11.003

G Model CATTOD-8717; No. of Pages 9 6

ARTICLE IN PRESS no et al. / Catalysis Today xxx (2013) xxx–xxx M. Maro˜

Fig. 7. Permeation of hydrogen in N2 /CO/H2 ternary mixtures: effect of carbon monoxide on H2 permeate flow rate at 450 ◦ C.

Experimental data obtained suggested that low temperatures (320 ◦ C) should be avoided for using the membrane in presence of high carbon monoxide partial pressures. Regarding the other syngas components in the feed gas stream, the presence of steam seemed to have the strongest negative influence in the permeation of H2 due to the concentration–polarization effect [36]. Due to the rather high steam contents required in the process, detailed studies are on-going to quantity the effect of steam on the permeation of hydrogen in the new membrane reactor configuration. Despite the loss in permeate capacity, no indication of the occurrence of secondary non-desired reactions was found. The relatively low molar ratio H2 /CO used in the feed gas mixture guaranteed that no methanation or disproportionation reactions took place and only H2 , N2 and CO was measured in the retentate stream. Based on the results obtained up to now, temperatures above 350 ◦ C are recommended for a correct performance of the membrane under conditions of SEWGSMR and although the inside–outside flux provides a slightly lower permeation that the outside–inside configuration (see Fig. 5), the membrane has showed an adequate performance for their use in the hybrid system proposed. 3.4. Performance of the binary system adsorbent/catalyst

Fig. 8. Permeation of hydrogen in N2 /CO/H2 ternary mixtures: effect of temperature on hydrogen permeation.

influence on the flux of hydrogen that permeates through the membrane when compared to pure H2 but the variations in the flux of H2 permeated when different concentration of CO in the feed gas are fed are not so significant. The other key parameter which has been reported in the literature to have strong influence on the performance of the membrane in the presence of CO was process temperature. A second series of experiments were carried out to evaluate the performance of the membrane in the temperature range of interest of the hybrid system (300–500 ◦ C). Fig. 8 shows the effect of temperature changes on the permeation of H2 when both pure H2 and a N2 /CO/H2 ternary mixture are used. As it can be observed, the drop in hydrogen permeation is particularly noticeable for pure H2 , losing approximately one third of the initial value when temperature is decreased from 735 to 598 K. In the case of the ternary mixture tested the effect is even more pronounced, losing approximately half of the initial H2 permeation flux. A plausible explanation for these differences in behaviour between pure H2 and ternary mixtures H2 /N2 /CO can be found in the inhibiting effects of co-existing gases in the permeation of H2 . According to the literature [39], as temperature decreases, CO is probably adsorbed on the surface of the membrane causing an additional barrier for H2 diffusion.

The advantages of combining a WGS catalyst and a high temperature CO2 sorbent focus on the process intensification concept, concentrating in a single reactor both processes (reaction and adsorption) and on the enhancing effect that the presence of the adsorbent has on the catalytic activity of the catalyst. Once the best operating conditions, in terms of CO conversion and CO2 capture, have been defined for the catalyst and the sorbent separately, it is necessary to optimise the performance of the combined system in order to take advantage of the possible synergies between them. One of the operating parameters that needs to be defined for the hybrid system is Vads /Vcat , that is, the relative volume of adsorbent (ads) and catalyst (cat) in the solid mixture. The adsorbent–catalyst mixture used in the proposed hybrid system should contain enough catalyst to provide sufficient conversion of CO and enough adsorbent to capture as much CO2 as possible at the selected process temperature. In our case we used two different volume ratios Vads /Vcat (5 and 10) and we performed the tests at temperatures in the range of 300–350 ◦ C and 15 bar using the same excess of steam (50%, v/v). According to the results obtained (data not showed, see Ref. [40] for a detailed description), complete conversion of CO was found for the first sorption cycle for both volume ratios adsorbent/catalyst tested, most likely due to the catalytic activity showed by the adsorbent. Therefore, a (Vads /Vcat ) = 10 was selected for the hybrid adsorbent/catalyst tests. Based on the individual performance showed by the catalyst and the sorbent versus temperature (see respectively Figs. 2 and 3), two experiments have been carried out for the adsorbent/catalyst mixture: the first one at 300 ◦ C using an excess steam of 50% and another one at 375 ◦ C using an excess steam of 23%. These two operating conditions were selected to investigate the combined effect of temperature and steam on the global performance of the binary system adsorbent/system before its integration in the hybrid adsorbent/catalyst/membrane. The steam percentage of 23% (v/v) corresponds to a steam to carbon monoxide molar ratio of 3, which is a very common value, used in WGS processes to guarantee that secondary reactions do not take place during the reaction. Different mechanisms have been proposed in the literature for CO2 capture using hydrotalcite based materials including the reconstruction of the hydrotalcite structure or the formation of carbonates and, in general, all these processes require the presence of steam, as reactant or just as an intermediate in the

˜ et al., Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 Please cite this article in press as: M. Marono, production under pre-combustion operating conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.11.003

G Model CATTOD-8717; No. of Pages 9

ARTICLE IN PRESS no et al. / Catalysis Today xxx (2013) xxx–xxx M. Maro˜

Fig. 9. Cyclic performance of binary system adsorbent/catalyst. P = 15 bar, T = 300 ◦ C, 50% steam.

capture reactions. This means that during the CO2 capture process consumption of steam will occur reducing the excess of steam available for the WGS. Besides that, the sorbent itself adsorbs water, even at high temperature, and this amount of water should also be taken into account during the sorption enhanced WGS process. Both experiments were performed using a maximum feed partial pressure of CO around 0.7 bar. Again, four sorption–desorption cycles were performed. In both cases desorption was carried out at 1 bar and 500 ◦ C using N2 as desorption agent. According to the results obtained for the adsorption process at 300 ◦ C (shown in Fig. 9), an anticipated breakthrough of CO took place reaching a final CO conversion of 40–45% after the fourth adsorption cycle. Despite the rather high amount of steam available (50%, v/v) the temperature was too low to guarantee simultaneous breakthrough of CO and CO2 . This behaviour suggested that when the SEWGS process is performed at 300 ◦ C the catalytic activity showed by the mixture adsorbent + catalyst was the sum of the individual catalytic activity of both solids. Under these operating conditions an increase in the amount of catalyst used seemed to be necessary. Besides carbon dioxide, H2 is also captured during the sorption step and the excess of steam fed was enough to guarantee that side reactions did not occur (methane was not detected in the exit gas stream during the whole test). When the SEWGS process was carried out at 375 ◦ C complete conversion of CO was expected to occur based on the performance showed previously by the adsorbent alone at temperatures higher than 350 ◦ C (see Fig. 4). However, as it can be seen in Fig. 10, an anticipated breakthrough of CO took place again although the final conversion of CO reached values higher than 90%. In this case a lower steam contents was used during the first three sorption cycles of the test (23%, v/v steam) and the results obtained suggested that it was not enough to convert all CO. The competition for the available steam during the sorption enhanced WGS process might justify the early breakthrough of CO during the three first sorption cycles shown in Fig. 10. To check if more steam was required, during the fourth adsorption cycle the amount of steam was increased to a value similar to that used previously (41%, v/v) and as it can be observed in Fig. 10 the conversion obtained for CO was almost 100% and the selectivity to H2 and CO2 was close to one. During those tests no indication of the presence of secondary non-desired reactions was found

7

Fig. 10. Cyclic performance of binary system adsorbent/catalyst. P = 7.5 bar, T = 375 ◦ C, R(H2 O/CO) = 3–7. Feed gas composition: 9% (v/v) CO/N2 .

by analysis of the exit gas stream. Only H2 , N2 , CO and CO2 were measured, guaranteeing that methane formation did not occur. This behaviour suggested that due to the catalytic activity of the sorbent, process temperatures higher than 350 ◦ C are beneficial during the sorption stage but due to the competitive avidity for water between the WGS reaction and the CO2 capture process the use of an excess steam higher than R = 3 is recommended to obtain a complete conversion of CO. The complete conversion of CO obtained during the sorption step before the breakthrough of CO2 is due to the sorption-enhanced effect of the capture of CO2 . 3.5. Hybrid adsorbent–catalyst–membrane system The main objective of the proposed hybrid system is to investigate the advantages of combining a SEWGS reactor for CO2 capture and a membrane reactor for H2 production and separation in one single process. Both the adsorbent and the catalyst were crushed and sieved to a particle size between 1.6 and 2.3 mm to favour homogeneity of the solid phase in the reactor. As mentioned above, the solid was located in the inner part of the membrane tube. SiC was used as inert material filling the membrane tube up to the beginning of the membrane area. Then the prepared mixture of adsorbent and catalyst was added until the other extreme of the membrane was reached. A total amount to 47 g of adsorbent/catalyst mixture with a volume ratio Vads /Vcat = 10 was used in the experiments. The operating conditions defined for this experiment were selected based on the individual performance of the membrane and the adsorbent/catalyst system described previously in this paper and the maximum operating conditions available for the membrane reactor facility. Thus, a feed gas composition which consisted of a mixture of 10% CO/N2 was used and the test was carried out at 360–380 ◦ C and at 7.5 bar in the retentate side and atmospheric pressure in the permeate side. No sweep gas was used. The amount of steam used in this experiment was 23% (v/v) which represents a molar steam to CO ratio of 3. Despite the requirements of steam observed in the experiments performed with the adsorbent–catalyst system, this value was selected to investigate if the simultaneous removal of H2 during the SEWGS process provided the complementary enhancement of the WGS reaction to obtain complete conversion of CO before the breakthrough of CO2 (in comparison with results presented in Fig. 10).

˜ et al., Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 Please cite this article in press as: M. Marono, production under pre-combustion operating conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.11.003

G Model CATTOD-8717; No. of Pages 9 8

ARTICLE IN PRESS no et al. / Catalysis Today xxx (2013) xxx–xxx M. Maro˜

shift reaction might be taking place. This phenomenon has already been reported to occur in presence of palladium-based membranes by Goldbach and co-workers [41]. Besides that when we opened the membrane reactor the presence of soot was evident showing that the Boudouard reaction might have happened to some extent perhaps during the sorption process due to the competing avidity for steam between the sorbent and the catalyst (during this experiment a steam to CO molar ratio of 3 was used). Later, after breakthrough of CO2 this soot might have reacted to produce CO contributing to the increase of the CO measured in the retentate side. Based on the results obtained, the advantages of using a membrane in combination with a sorption enhanced water gas shift reaction seem to be positive because complete conversion of CO is obtained up to breakthrough of CO2 and a pure stream of H2 can be produced in the permeate side. However, in order to avoid the occurrence of the reverse WGS reaction on the membrane surface additional excess of steam is confirmed to be needed. Fig. 11. Gas composition profiles in the retentate side during second adsorption cycle of the hybrid adsorbent–catalyst–membrane experiment. P = 7.5 bar, Feed: 10% CO/N2, 23% steam.

Then, two sorption–desorption cycles were carried out and during this experiment regeneration was performed using a mixture of steam and nitrogen (23%, v/v steam) at 500 ◦ C and 1 bar. Under those conditions a simultaneous breakthrough of CO and CO2 and an increase in the conversion of CO are expected to occur due to the permeation of H2 . However, the presence of the membrane changed the foreseen performance of the system. Both sorption cycles described very similar breakthrough profiles: almost simultaneous breakthrough of CO and CO2 took place providing a conversion of CO higher than 95% (v/v) but after breakthrough of CO2 the conversion of CO started to decrease smoothly reaching conversion values around 20% (v/v) which was much lower than the measured for the catalyst or the sorbent individually at this temperature. Fig. 11 shows the exit gas composition of the retentate side obtained for the second adsorption cycle. As it can be observed in Fig. 11, as expected, almost simultaneous breakthrough of CO and CO2 took place and almost complete removal of H2 took place. Up to breakthrough only a slipstream of H2 in N2 was measured in the retentate side while the main amount of the produced H2 was supposed to be present in the permeate stream. However, no permeation of hydrogen was measured. Although we do not have a clear explanation for this, we suggest two possible situations: (1) that H2 could be permeating but we could not measure it due to the very low flux of hydrogen produced or (2) no permeation took place because the driving force was not enough to allow H2 permeation under the conditions tested. In this case the produced hydrogen could be retained inside the membrane or captured by the adsorbent somehow. Whichever the mechanism for H2 removal in this experiment was, the effect of the removal of H2 was positive for the WGS reaction, increasing the CO conversion up to 100% up to CO2 breakthrough. Another very interesting difference is the shape of the breakthrough curves. In general, in the adsorbent–catalyst experiments performed up to now, a stable final CO2 and CO composition is reached after the breakthrough of CO and the final value of the conversion of CO depended on the process operating conditions, i.e. temperature, steam contents or feed gas composition. However, in this experiment after breakthrough of CO2 the concentration of CO2 in the retentate stream abruptly decreased in coincidence with a decrease in the conversion of CO and a complete removal of H2 . This behaviour suggested that after saturation of the sorbent at 375 ◦ C it started to release CO2 which in presence of H2 (retained in the sorbent or within the membrane) was consumed to produce CO, that is, under those operating conditions the reverse water gas

4. Conclusions In this work the performance of a hybrid system adsorbent–catalyst–membrane is studied. First the individual performance of the catalyst and the sorbent previously selected were tested. Both materials were combined in a single reactor and the operating conditions for this binary system were optimised in terms of CO conversion and CO2 capture. Finally a Pd selective membrane was added to this catalyst–adsorbent mixture and a novel adsorbent–catalyst–membrane system was designed. From the results obtained during this work the following conclusions can be drawn: - A compromise is needed when selecting the process temperature for the binary system catalyst/sorbent used in this work: low temperatures favoured the CO2 capture process while high temperatures favoured the conversion of CO. In our case a value in the range of 350–375 ◦ C was found to be adequate. - Depending on the process temperature selected a different volume ratio Vads /Vcat might be needed. Specific studies must be done for each catalyst–adsorbent system used. - An additional excess of steam is required in the sorption enhanced WGS process to guarantee the adequate performance of the hybrid system in terms of complete conversion of CO. - The presence of the membrane has positive effects on the catalytic behaviour of the hybrid system up to CO2 breakthrough. - However, detailed studies must be done to optimise the operating conditions in order to avoid the occurrence of the reverse water gas shift reaction due to the presence of the palladium membrane. - The increase in the excess steam available during the process is proposed as a plausible solution and detailed studies will be performed to confirm this hypothesis. - Specific studies must be also done for clarification of the real effects occurring during the combined processes. Acknowledgements Authors wish to thank the Spanish Ministry of Science and Innovation (ENE2009-08002 CAPHIGAS project) and the European Community (RFCS-CT-2010-00009 FECUNDUS project) for their financial support. We also thank the reviewers for their comments that have contributed to improve the quality of this paper. References [1] H.A.J. van Dijk, S. Walspruher, P.D. Cobden, R.W. van den Brink, F.G. de Vos, Int. J. Green House Control 5 (2011) 505–511.

˜ et al., Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 Please cite this article in press as: M. Marono, production under pre-combustion operating conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.11.003

G Model CATTOD-8717; No. of Pages 9

ARTICLE IN PRESS no et al. / Catalysis Today xxx (2013) xxx–xxx M. Maro˜

[2] N. MacDowell, N. Florin, A. Buchard, J. Hallett, A. Galindo, G. Jackson, C.S. Adjiman, C.K. Williams, P. Fennell, Energy Environ. Sci. 3 (2010) 1645–1669. [3] Q. Wang, J. Luo, Z. Zhong, A. Borgna, Energy Environ. Sci. 4 (2011) 42–55. [4] P. Luis, T. van Gerven, B. van der Bruggen, Prog. Energy Combust. Sci. 38 (2012) 419–448. [5] F. Gallucci, E. Fernández, P. Corengia, M. van Sint Annaland, Chem. Eng. Sci. 92 (2013) 40–66. [6] A. Brunetti, G. Barbieri, E. Drioli, K.H. Lee, B. Sea, D.W. Lee, Chem. Eng. Proc. 46 (2007) 119–126. [7] T.A. Peters, M. Stange, H. Klette, R. Bredesen, J. Membr. Sci. 316 (2008) 119–127. ˜ [8] J.M. Sánchez, M. Barreiro, M. Marono, Biomass Bioenergy 35 (2011) S132–S144. [9] Z. Yong, E. Rodriges, Energy Convers. Manage. 43 (2002) 1865–1876. [10] P.D. Cobden, et al., Int. J. Greenhouse Gas Control 1 (2007) 170–179. [11] Y. Ding, E. Alpay, Chem. Eng. Sci. 55 (2000) 3461–3474. [12] R.J. Allam, R. Chiang, J.R. Hufton, P. Middleton, E.L. Weist, V. White, in: D.C. Thomas, S.M. Benson (Eds.), Carbon Dioxide Capture for Storage in Deep Geologic Formations, vol. 1, Elsevier Ltd, 2005, pp. 227–256. [13] J.R. Hufton, S. Mayorga, S. Sircar, Separations 45 (2) (1999) 248–256. [14] A. Iwan, A. Lapkin, Proceedings of International Hydrogen Energy Congress and Exhibition IHEC 2005, Istanbul, Turkey, 13–15 July, 2005. [15] H.Th.J. Reijers, S.E.A. Valster-Schiermeier, P.D. Cobden, R.W. van den Brink, ECNRX-05-122, 2005. [16] E. van Dijk, S. Walspurger, P. Cobden, R. van den Brink, Energy Proc. 4 (2011) 1110–1117. [17] A. Harale, H.T. Hwang, P.T.K. Liu, M. Sahimi, T. Tsotsis, Chem. Eng. Sci. 62 (2007) 4126–4137. ˜ [18] M. Marono, Y. Torreiro, L. Montenegro, J.M. Sánchez, Fuel (2013), doi:10.16/j.fuel.2013.03.067. ˜ J.M. Sánchez, E. Ruiz, Int J. Hydrogen Energy 35 (2010) 37–45. [19] M. Marono, ˜ D. Cillero, L. Montenegro, E. Ruiz, Fuel (2012), [20] J.M. Sánchez, M. Marono, doi:10.16/j.fuel.2012.02.060. ˜ [21] J.M. Sánchez, M.M. Barreiro, M. Marono, Fuel (2013), doi:10.16/j.fuel.2013.02.051. ˜ P. Casero, J. Trujillo, Proceedings of the 5th International Freiberg Con[22] F. Pena, ference on IGCC& XtL Technologies, Leipzig, 21–24 May, 2012, p. 62, Book of Abstract. [23] KATALCOJM, http://jmcatalysts.com/ptd/pdfs-uploaded.

9

˜ Y. Torreiro, L. Gutierrez, Int. J. Greenhouse Gas Control 14 (2013) [24] M. Marono, 183–192. [25] A.D. Ebner, S.P. Reynolds, J. Ritter, Ind. Eng. Chem. Res. 45 (2006) 6387–6392. [26] S. Walspurger, L. Boels, P.D. Cobden, G.D. Elzinga, W.G. Haije, R.W. van den Brink, ChemSusChem 1 (2008) 643–650. [27] E. Xue, M. Okeeffee, J.R.H. Ross, Catal. Today 30 (1996) 107–118. [28] A. Effendi, K. Hellgardt, Z-G. Zhang, T. Yoshida, Fuel 84 (2005) 869–874. [29] S. Walspurger, P.D. Cobden, O.V. Safonova, Y. Yu, E.J. Anthony, Chem. Eur. J. 16 (2010) 12694–12700. [30] M.K. Ram Reddy, Z.P. Xu, G.Q. Lu, J.C. Diniz de Costa, Ind. Eng. Chem. Res. 47 (2008) 2630–2635. [31] S.C. Lee, B.Y. Choi, C.K. Ryu, Y.S. Ahn, T.J. Lee, J.C. Kim, Korean J. Chem. Eng. 23 (3) (2006) 374–379. [32] E.R. van Selow, P.D. Cobden, P.A. Verbraeken, J.R. Hufton, R.W. van den Brink, Ind. Eng. Chem. Res. 48 (2009) 4184–4193. [33] S. Tosti, M. Zerbo, A. Basile, V. Calabrio, F. Borgognoni, A. Santurcci, Int. J. Hydrogen Energy 38 (2013) 701–707. ˜ J.M. Sánchez, Proceedings of the 6th International [34] M.M. Barreiro, M. Marono, Conference on Clean Coal Technologies 2013, Tessaloniki, 12–16 May, 2013, p. 146, Book of Abstracts. [35] O. Iyoha, B. Howard, B. Morreale, R. Killmeyer, R. Enick, Top. Catal. 49 (2008) 97–107. [36] A. Unemoto, N. Hirai, A. Kaimai, K. Sato, T. Otake, K. Yashiro, J. Mizusaki, T. Kawada, T. Tsuneki, Y. Shirasaki, Y. Ota, I. Yasuda, Int J. Nucl. Hydrogen Prod. Appl. 2 (1) (2009) 55–68. [37] H. Li, J.W. Dijkstra, J.A.Z. Pieterse, J. Boon, R.W. van den Brink, D. Jansen, J. Membr. Sci. 363 (2010) 204–211. [38] T. Flanagan, D. Wang, K.L. Shanahan, Phys. Chem. Chem. Phys. 2 (2000) 4976–4982. [39] F. Scura, G. Barbieri, E. Drioli, Desalination 200 (1–3) (2006) 239–241. ˜ Y. Torreiro, J.M. Sánchez, Proceedings of the 6th International Con[40] M. Marono, ference on Clean Coal Technologies 2013, Tessaloniki, 12–16 May, 2013, p. 148, Book of Abstracts. [41] H. Li, A. Goldbach, C.H. Tang, J. Boon, J.A.Z. Pieterse, J.W. Dijkstra, R.W. van den Brink, C. Bao, H.Y. Hu, Proceedings of the International Conference on Catalysis in membrane reactors ICCMR 10, St. Petersbourg, June 2010, 2010, p. 34, Book of Abstracts.

˜ et al., Performance of a hybrid system sorbent–catalyst–membrane for CO2 capture and H2 Please cite this article in press as: M. Marono, production under pre-combustion operating conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.11.003