Correlation between structural rearrangement of hydrotalcite ... - Core

Energy Procedia 4 (2011) 1162–1167 Energy Procedia 00 (2010) 000–000

Energy Procedia www.elsevier.com/locate/procedia www.elsevier.com/locate/XXX

GHGT-10

Correlation between structural rearrangement of hydrotalcite-type materials and CO2 sorption processes under pre-combustion decarbonisation conditions. S. Walspurger, a*1 S. de Munck,a P.D. Cobden,a W.G. Haije,a R.W. van den Brink,a O.V. Safonovab b

a Energy research Centre of the Netherlands (ECN), Westerduinweg 3, 1755ZG, Petten, Netherland Swiss-Norwegian Beam Lines (SNBL) European Synchrotron Radiation Facility (ESRF), 6 Rue Jules Horowitz, 38043 Grenoble, France

Elsevier use only: Received date here; revised date here; accepted date here

Abstract Alkali-promoted hydrotalcite-based materials have showed CO2 sorption capacity up to 1.6mmol.g-1 when CO2 separation is carried out at relatively high temperature (300-500ºC) and high partial pressure of steam and CO2, conditions that are prevalent in pre-combustion CO2 capture conditions. In addition to CO2 breakthrough experiments in situ XRD experiments have allowed identifying magnesium carbonate crystalline phase formation during CO 2 sorption. The alkali promoted hydrotalcite-based material (Mg/Al ratio 2.9) has revealed that high CO2 capacities can potentially be achieved by using the extra storage capacity offered by the presence of excess magnesium oxide that can be converted into magnesium carbonate. The use of this extra capacity in Pressure Swing Adsorption process is further discussed. © 2010 Elsevier Ltd. All rights reserved c 2011 Published by Elsevier Ltd. Open access under CC BY-NC-ND license. ⃝ Keywords: sorption enhanced reaction; carbon capture; XRD; pressure swing adsorption; hydrotalcite.

1. Introduction

Greenhouse gas mitigation may be achieved in electricity generation processes by implementing pre-combustion decarbonisation in coal gasification or natural gas reformate fired power plants. Reducing the efficiency penalty induced by CO2 removal is an essential target that may be reached by implementing novel concepts leading to more efficient separation technologies. Sorption Enhanced Water-Gas Shift (SEWGS) process which has been developed since the late 90’s,[1] provides a smart way to simultaneously convert CO in syngas to CO 2 and H2 and remove CO2. The resulting hot and pressurized H2-rich gas stream may be combusted in a gas turbine to generate electricity. SEWGS reaction is typically carried out in a Pressure Swing Adsorption (PSA) unit at temperature around 400°C and pressures of about 30-50 bar depending on the syngas generation technology used. Once the sorbent materials used for CO2 adsorption has been saturated, CO2 is recovered by regeneration with a purge of low pressure superheated steam (about 400ºC). To ensure continuous production of H2 the PSA unit is equipped with multiple * Corresponding author. Tel.: +31-224-564116; E-mail address: [email protected].

doi:10.1016/j.egypro.2011.01.169

2

S.Walspurger Energy Procedia 00 (2010) 000–000 S. Walspurger et al.et/ al./ Energy Procedia 4 (2011) 1162–1167

1163

SEWGS reactors interconnected with each other such it may operate in cyclic mode.[2] In the research and development of such an industrial gas separation process, it is of utmost importance to understand the behaviour of the sorbent material under relevant conditions of pressure, temperature and gas mixture to further optimize its characteristics accordingly to the constraints and targets of the industrial process (lifetime, mechanical strength, CO 2 capacity, resistance towards impurities). Alkali promoted hydrotalcite-based materials (Aluminium-Magnesium containing clays) have been found to work very well under SEWGS conditions showing fair working capacity and appropriate kinetics of sorption.[1-3] In parallel with our bench scale demonstration units, mechanistic studies on CO2 adsorption have been carried out to identify structural rearrangement and active sorption sites in sorbent materials. Although quite a lot of data are available in the literature for characterization at atmospheric pressure and higher pressure under dry conditions,[4-6] experimental data at relevant pressure and presence of steam are essential for sorbent development and further adsorption modelling work. This study aimed at correlating CO2 adsorption experiments at about 300-500°C in the presence of high partial pressure steam with corresponding chemical and structural rearrangement occurring in the sorbent material.

2. Experimental For the measurement of CO2 adsorption capacity, a 500mm tubular reactor with an inner diameter of 13.4mm was loaded with 31.0g of 20wt% potassium carbonate promoted hydrotalcite based material (Mg/Al = 2.9) provided by Sasol Germany which was located between two beds of low surface area alpha-alumina (NorPro SA 52124 1.3 mm spheres) used as bed support material. The sorbent consisted of sieve fraction 0.425-0.212mm. The test rig was equipped with an infrared detector Hartmann and Braun with 4 cells for dry gas composition analysis : Uras 26 for CO 0-10%, CH4 0-10% and CO2 0-30%, and Caldos 17 for H2:N2 0-100%. The sorbent activation procedure was carried out by heating in dry N2 (1bar) at a total flow of 450ml/min-1 to 450ºC for 5h followed by a temperature decrease to 300ºC. The pressure was then set to 20bar and steam was introduced using a steam generator at a flow of 6.82g.h-1. After stabilisation of the detector signal, the gas composition was changed for adsorption experiment such that CO2, steam and (methane tracer +) N2 reached 5bar, 5bar and 10bar respectively with a total flow of 600ml.min1 (STP). Long adsorption step presented in figure 1 was carried out for 250min with preliminary activation of the sorbent for 24h with steam and N2 at 400ºC. In situ powder XRD data were recorded at the SNBL (Swiss-Norwegian Beam Lines, BM01B) of the 6 GeV ESRF synchrotron (Grenoble, France) as described in a previous report.[7] High pressure CO2 and steam could be fed into a gas cell that has been designed such that no water condensation could occur on the sample.

3. Results and discussion It is well known that hydrotalcite-based materials are converted in a mixed metal oxide upon heating between 300 and 500°C.[8] Dehydroxylation and decarbonation above 350°C lead to a crystalline mixed oxide showing broad diffraction peaks corresponding to periclase (or salt rock) crystalline phase generally described as Mg(Al)O. For potassium carbonate promoted hydrotalcite-based materials a similar decomposition pattern has been observed under atmospheric pressure.[9] While non-promoted mixed oxide Mg(Al)O showed only very low reversible CO2 adsorption capacity at 350-450ºC, potassium carbonate promoted corresponding materials showed good performances in cyclic CO2 adsorption/desorption at these temperatures under ordinary pressure. Only few reports on sorption experimental data carried out at relevant pressure conditions for industrial applications can be found in the literature. None of them are reporting sorption data in the presence of high partial pressure steam (which is typically associated with Water-Gas Shift conditions) mainly because achieving stable conditions with such humid conditions is not trivial and demands rigorous heat tracing of test-rigs and suitable online analysis. In our effort to understand how the sorbent material works and behaves under conditions that are as representative as possible for SEWGS conditions, we have tested the CO2 sorption properties of the material in a dedicated test-rig on the one hand and have characterised by high resolution XRD on the other hand using comparable conditions. Figure 1 reports CO2 breakthrough curves measured under total pressure of 20 bar with 5 bar pressure of steam and 5bar

1164

Procedia 00 4(2010) 000–000 S.S.Walspurger Walspurgeretetal./ al.Energy / Energy Procedia (2011) 1162–1167

3

partial pressure of CO2 at 350ºC. The sorbent had been regenerated for 24h with steam and N 2 at 400ºC before extended CO2 breakthrough experiment started.

400

30

390 25

380

CO2

370

temperature 360

15

350 340

10

Bed temperature (ºC)

Molar compostion (%)

Inert tracer 20

330

320

5

310 300

0 0

20

40

60

80

100

120

140

Adsorption time (min)

Figure 1 CO2 breakthrough experiment at 350ºC with a total pressure of 20 bar, 5 bar of steam, 5 bar of CO2 and 10 bar of N2+inert. Temperature profile in the middle of the absorber bed is inserted on the y-scale on the left. During the adsorption time the inert gas concentration rapidly increased to a level that exceeded its feed concentration because of the CO2 adsorption (total molar flow decreased while molar flow of inert tracer remained constant). The breakthrough time corresponded to a breakthrough capacity of 1.6mmol.g -1, which is comparable to the results reported earlier and is quite high compared to literature values mostly because of the extended regeneration time that preceded the adsorption under these pressure conditions. Remarkably after CO2 breakthrough the CO2 concentration did not immediately reach the feed concentration (about 26%), but showed a wave that slowly reached the feed level instead. This behaviour had previously been observed for such materials (20wt% potassium carbonate promoted hydrotalcite based material with Mg/Al=2.9) in SEWGS conditions.[2] Furthermore the bed temperature profile clearly exhibited a sharp exotherm followed by a more diffuse exotherm that would be related to the CO2 concentration wave reaching very slowly its feed concentration. Accordingly it is quite intuitive to correlate the first part of the breakthrough curve with a fast carbonation reaction while the second part would result from a slower process associated with bulk carbonate formation upon exposure to high pressure CO 2. Very similarly results reported earlier on CO2 adsorption using different gas velocities with similar gas composition at 400ºC [2,3] indicated that the sorbent keeps taking up rather significant amount of CO2 after breakthrough as a result of a bulk carbonate formation. In-situ characterization studies have been carried out to elucidate the chemical process that is responsible for such a sorption behaviour. Therefore we have designed a test-rig that allows reaching such conditions of temperature and pressure in the presence of high steam partial pressure in collaboration with the Swiss Norwegian Beam Lines at the European Synchrotron Radiation Facility (Grenoble, France) by using the in-house gas rig and microreactor holder depicted in a previous report.[7]

4


1165

Figure 2 In situ XRD analysis of 20wt%K2CO3 promoted hydrotalcite based material at 400ºC and under 1) 5 bar of steam, 5 bar of CO2, 2) 1 bar dry N2, 3) increased dry CO 2 pressure from 1 to 10 bar, 4) increased steam partial pressure from 0 to 5bar in 10 bar pressure balanced with CO 2. When the potassium carbonate promoted hydrotalcite based material was contacted at 400°C with a gas mixture consisting of 5 bar of CO2 and 5 bar of steam a set diffraction peaks corresponding to crystalline magnesium carbonate clearly appeared as a consequence of carbonation reaction of magnesium (hydr-)oxide centres (step 1 on figure 2). While mixed oxide Mg(Al)O periclase crystalline phase can be usually observed at ordinary pressure, figure 2 clearly shows that Mg(Al)O mixed oxide underwent carbonation which led to formation of crystalline MgCO3. The experiment was further carried on by simulating a pressure swing cycle which consisted of first decreasing the pressure from 10 to 1bar and switching the gas composition from a 1:1 CO2:H2O to a 100% dry N2 while maintaining the temperature at 400ºC. As a result MgCO3 crystalline phase completely disappeared while MgO periclase crystalline phase diffraction peaks became visible. It is worthwhile noticing that the diffraction peaks observed here were much narrower than the ones usually observed upon thermal decomposition of hydrotalcite.[8] This may be a direct consequence of precedent magnesium carbonate formation and decomposition that could have led to formation of mostly Aluminium-defect free MgO periclase structure as opposed to the periclase-like structure directly derived from hydrotalcite thermal decomposition where Aluminium centres are still present in the distorted cubic salt-rock like structure. Moreover it is worthwhile mentioning that crystalline γ-alumina-type phase is visible which may result from a phase segregation between Mg and Al that has been enhanced by the carbonationdecarbonation process. The experiment was further carried out by switching the gas mixture from N2 to CO2 (dry) and increasing the total pressure to 10 bar. The XRD pattern recorded during the pressure increase showed no major structural changes. When a pressure of 10 bar dry CO 2 was reached, steam was introduced with a relative concentration 1:1 towards CO2. XRD patterns at the top of figure 2 (step 4 on the graph) show the evolution of the

1166

Procedia 00 4(2010) 000–000 S.S.Walspurger Walspurgeretetal./ al.Energy / Energy Procedia (2011) 1162–1167

5

structure as a function of time following the introduction of steam: the first from the bottom is taken after 10 minutes addition of steam, the second one after 20 minutes and the top one after 30 minutes. In the presence of steam, magnesium carbonate crystalline phase re-appeared progressively, while the magnesium oxide periclase phase gradually disappeared. These observations are well in agreement with the thermodynamic data on magnesium oxide, magnesium hydroxide and magnesium carbonate systems. The increase of steam pressure may induce a partial rehydroxylation of MgO periclase phase which enhances the formation of magnesium carbonate[10,11]. Obviously K2CO3 promoted hydrotalcite based material studied in the present work can reversibly form crystalline magnesium carbonate in the presence of steam. However it appeared clearly in Figure 1 and previous results that only a small part of the corresponding CO2 capacity may actually participate in the pre-breakthrough region in which most of the CO2 capacity comes from the formation of potassium(hydr-)oxyalumino carbonate species (KO(H)-Al-CO3) [9] that is probably a faster and more reversible process than bulk magnesium carbonate which may be severely limited by mass transport through the carbonate layer. Although bulk carbonate formation may bring larger CO2 storage capacity to the sorbent, it may severely affect both the stability of the sorbent as well as the capture ratio in CO 2 separation by Pressure Swing Adsorption. The reversible conversion of magnesium oxide to magnesium carbonate indeed implies back and forth structural rearrangement that may eventually alter the mechanical integrity of the material if thorough regeneration is carried out [12]. Moreover the presence of remaining bulk carbonate in the reactor (when using reasonable regeneration time and conditions) having slower carbonation-decarbonation dynamics than surface carbonates may induce a significant slip of CO2 in the H2 product during adsorption which may alter the separation performance.[2] For these reasons excessive bulk carbonate is not desirable and ongoing work on sorbent development for SEWGS focuses on reaching the optimal trade-off between sorption reversibility and CO2 capacity. 4. Conclusion CO2 adsorption experiments carried out on alkali-promoted hydrotalcite-based materials (Mg/Al ratio of 2.9) under conditions of pressure and temperature relevant to Sorption-Enhanced Water-Gas Shift reaction have shown that relatively high CO2 capacity could be achieved depending mostly on the regeneration conditions that are used before the adsorption experiment. Magnesium carbonate formation has been identified by in situ XRD as an important process taking place during CO2 adsorption step. Most of the extra CO2 capacity that is induced by magnesium carbonate formation upon long adsorption experiment is however hardly beneficial for SEWGS process due to the highly dynamic nature of pressure swing cycles compared to the magnesium carbonate formation and decomposition. Ongoing material development currently focuses on reaching the optimal trade-off between sorption reversibility and CO2 capacity.

5. Acknowledgment The authors thank the Dutch Ministry of economic affairs for financial support. NWO (Dutch organization for scientific research) and Wim Bras, manager of DUBBLE beam lines at ESRF are gratefully acknowledged for beam time allowance and fruitful technical exchanges. SW and PC would like to thank Gerard Elzinga and Dr. Ruud Westerwaal for their advices. 6. References - [1] Allam RJ, Chiang R, Hufton JR, Middleton P, Weist EL, White V, Development of the Sorption Enhanced Water-Gas Shift Process in Carbon dioxide capture for storage in deep geologic formations. Ed. D.C. Thomas and S.M. Benson, Elsevier Ltd, Oxford, UK, 2005 - [2] Van Selow ER, Cobden PD, van den Brink RW, Hufton JR, Wright A. Performance of sorption-enhanced water-gas shift as a pre-combustion CO2 capture technology. Proc. GHGT-9, Washington 2008. - [3] Van Selow ER, Cobden PD, Verbraeken PA, Hufton JR, van den Brink RW. Carbon capture by sorptionenhanced water−gas shift reaction process using hydrotalcite-based material. Ind Eng Chem Res 2009; 48: 4184-93.

6


1167

- [4] Kim Y, Yang W, Liu PKT, Sahimi M, Tsotsis TT. Thermal Evolution of the Structure of a Mg-Al-CO3 Layered Double Hydroxide: Sorption Reversibility Aspects. Ind Eng Chem Res 2004; 43: 4559-70. - [5] Lee KB, Verdooren A, Caram HS, Sircar, S. Chemisorption of carbon dioxide on potassium-carbonatepromoted hydrotalcite. J Colloid Interf Sci 2007, 308, 30-9. - [6] Du H, Williams CT, Ebner AD, and Ritter JA. In Situ FTIR Spectroscopic Analysis of Carbonate Transformations during Adsorption and Desorption of CO2 in K-Promoted HTlc. Chem Mater 2010; 22: 3519-26. - [7] Walspurger S, Cobden PD, Haije WG, Westerwaal R, Elzinga GD, Safonova OV. In Situ XRD Detection of Reversible Dawsonite Formation on Alkali Promoted Alumina: A Cheap Sorbent for CO2 Capture. Eur J Inorg Chem 2010; 2461-64. - [8] Debecker DP, Gaigneaux EM, Busca G. Exploring, Tuning, and Exploiting the Basicity of Hydrotalcites for Applications in Heterogeneous Catalysis. Chem Eur J 2009; 15: 3920-35. - [9] Walspurger S, Boels L, Cobden PD, Elzinga GD, Haije WG, van den Brink RW. The crucial role of the K+aluminium oxide interaction in K +-promoted alumina- and hydrotalcite-based materials for CO2 sorption at high temperatures. ChemSusChem 2008; 1: 643-50. - [10] Siriwardane RV, Stevens Jr RW. Novel Regenerable Magnesium Hydroxide Sorbents for CO2 Capture at Warm Gas Temperatures. Ind Eng Chem Res 2009; 48: 2135-41. - [11] Bearat H, McKelvy MJ, Chizmeshya AVG, Sharma R, Carpenter RW. Magnesium hydroxide dehydroxylation/carbonation reaction processes: implications for carbon dioxide mineral sequestration. J Am Ceram Soc 2002; 85: 742-8. - [12] van Selow ER, Cobden PD, Wright A, van den Brink RW, Jansen D. Improved sorbent for Sorption Enhanced Water-Gas Shift process. Energy Procedia, this volume.