Xingyin Gao,Minghui Qiu,Kaiyun Fu,Peng Xu,Xiangli Kong,Xianfu Chen,Yiqun Fan*

State Key Laboratory of Materials-Oriented Chemical Engineering,College of Chemical Engineering,Nanjing Tech University,No.5 Xin Mofan Road,Nanjing 210009,China

Keywords:Absorption Membrane contactor Ceramic membrane Hydrophilic Desulfurization

ABSTRACT Hydrophilic ceramic membranes would be potentialcandidates for membrane gas absorption ifthey could be applied to appropriate separation processes.This study highlights a novel concept for the practical implementation ofSO2 absorption in hydrophilic ceramic membrane that exhibits outstanding thermaland mechanicalstabilities.With this aim,we investigated experimentally the performance of SO2 absorption into aqueous sodium hydroxide(NaOH)solution in a hydrophilic alumina(Al2O3)membrane contactorin terms ofSO2 removalefficiency and SO2 mass transfer flux,and compared the performance with that in a hydrophobic one.A series of experiments were performed atvarious conditions overa NaOHconcentration range of0-1.0 mol·L-1,a liquid flowrate range of30-180 ml·min-1,a gas flow rate range of120-1000 ml·min-1,an inlet SO2 concentration range of400-2000 μl·L-1,and a temperature range of10-35 °C.Itwas found thatthe hydrophilic membrane was more competitive when using a NaOH concentration higher than 0.2 mol·L-1.Furthermore,it can be inferred that the hydrophilic α-Al2O3 membrane exhibited exceptional long-term stability under 480 h continuous operation.

1.Introduction

Sulfur dioxide(SO2),a major air pollutant,is mainly produced by combustion of fossil fuels in power plants,incinerators,boilers and ships.Residual SO2in tail gas will not only contaminate air but also affect the subsequent CO2capture for it would react with amine-based solvents to form heat-stable salts which gradually lose CO2absorption capacity[1].At present,various technologies have been proposed to remove acid gas such as chemical and physical absorption,solid adsorption,cryogenic distillation,and membrane technology[2-4].

Membrane gas absorption(MGA)technology has attracted attention in the last three decades.Compared with conventional absorption devices,membrane contactors have many significant advantages such as operational flexibility,the independent gas and liquid flow,compact size,and scale-up[5].Currently,the majority of studies on MGA technology have largely focused on the CO2absorption with polymeric hollow fiber membrane(HFM)contactors,such as polypropylene(PP),polytetra fluoroethylene(PTFE)and polyvinylidene fluoride(PVDF),polyetheretherketone(PEEK),and polyethersulfone(PES)[6-10].However,efforts to investigate the membrane SO2absorption are far less.Relatively few studies have reported to investigate SO2absorption into various polymeric membranes such as PP,PVDF,and PTFE HFMs using seawater,pure water,and aqueous alkali solutions including NaOH,Na2SO3,Na2CO3,and NaHCO3[11-16].Compared with CO2,SO2is higher in reactivity and acidity while extremely lower in content.These properties ofSO2call forstable membrane materials to withstand chemical solvents for SO2absorption using MGA technology.

While the features of polymeric membranes including small diameter and thin-wall allows them to provide a high specific surface area for gasliquid contact,they suffer from weak anti-pollution,anti-chemical degradation,anti-thermalaging and mechanicalstrength.These drawbacks of polymeric membranes give rise to(i)pore blockage in the presence of suspended particles in flue gas;(ii)membrane damage in terms of morphology,microstructure and hydrophobicity;(iii)swelling and even liquid leakage after long-term exposure to chemical absorbent[17,18].

Compared with polymeric membranes,ceramic membranes show better structural,thermal,physical and chemical stability and have been widely applied in the separation and purification processes[19-21].In general,ceramic membranes for MGA tend to have asymmetric structures composed of a macroporous support layer,to ensure the mechanical strength of the overall membrane,and a thinner and denser membrane layer,to provide a fixed interfacial layer for gas and liquid contact and to prevent liquid to break through the membrane.Luis et al.[22,23]used a Al2O3hollow fiber membrane contactor to removalSO2from gas streams in N,N-dimethylaniline and discussed technical,environmentaland economic issues for industrialapplication.Han et al.[24]applied a single hydrophobic ZrO2ceramic tubule-based membrane contactor for SO2absorption with water as low-cost absorbent and found that the ceramic membrane contactor had lower height of transfer unit(HTU)value than conventional packed tower.Ionada Incorporated[25]applied semi-permeable ceramic hollow fibers to reduce emissions of target emission gases such as sulfur oxides,nitrogen oxides,and carbon oxides from marine engine exhaust.They reported that the Ionada Marine Scrubber is 50%smaller,and 30%more energy efficient than competitive salt water scrubbers.

According to whether the membrane pores are wetted by absorbent,the operating modes of the membrane contactor are mainly divided into non-wetted mode and wetted mode[5,26].The non-wetted mode,which refers to gas filled with membrane pores,is more widely studied.A general point is that hydrophobic membranes are more advantageous for gas absorption than hydrophilic ones,for gas absorption in nonwetted mode is many times higher than that in wetted mode[27,28].So far,most materials for MGA are hydrophobic or modified hydrophobic.Most ceramic membranes that are made of metal oxides,such as alumina(Al2O3),zirconia(ZrO2),silica(SiO2),are hydrophilic in nature with hydroxyl(--OH)groups on their surface.To operate MGA in a non-wetted mode,hydrophobic ceramic membranes are fabricated by grafting organosilane compounds with hydrophobic chains such as fluoroalkylsilanes(FASs)and chloroalkylsilanes on the surface of the hydrophilic ceramic membrane.For example,Yu et al.[29]used commercial ceramic membrane via grafting with FAS for CO2absorption.Lee et al.[30]applied the modified hydrophobic Al2O3HFM in a laboratory-scale gasliquid membrane contacting process for CO2capture applications using H2O as low-cost absorbent.Hydrophobic modification does enhance ceramic membrane performance for gas absorption,especially for those gases with high content,such as CO2,but it may not be the best for SO2.Karoor and Sirkar[11]reported that for CO2absorption in water,the wetted mode of operation offered considerably higher resistance to mass transferwhen compared to the non-wetted mode ofoperation.For SO2absorption in water,the wetted mode ofoperation offered somewhathigher resistance when compared with the non-wetted mode of operation.

Using a hydrophilic ceramic membrane for SO2absorption into aqueous solutionsunderwetted mode is a promising candidate as a relatively new approach for a number of reasons.Firstly,SO2is high in reactivity and extremely low in concentration,it can quickly react with absorbent while does not consume too much absorbent,which allows SO2to be absorbed into absorbent instantaneously.Secondly,hydrophobic modification of the ceramic membrane will increase cost.Moreover,even if the ceramic membrane has been modified,membrane wetting is still inevitable in a long-term operation.For example,Yu et al.[29]reported that the superhydrophobic ceramic membrane contactor was operated continuously for three weeks to evaluate its duration performance.The CO2removalefficiency continuously decreased with operating time and the membrane needed drying to ensure a high CO2removalefficiency.Lee and Park[31]used a porous alumina hollow fiber membrane which was modified with a FAS solution to test the long-term performance of CO2absorption.The CO2flux decreased by 15%from the initial value after 4360 min.Last but not least,α-Al2O3membranes show an excellent corrosion resistance[19,32]to ensure long-term stability of SO2absorption.

This study is to presenta comprehensive analysis of the potentialfor SO2absorption into water and aqueous NaOH solution using hydrophilic α-Al2O3ceramic membrane contactors.The performance was experimentally tested and compared with that of the hydrophobic one.In addition,the long-term stability was monitored to gain a better understanding of the system duration performance.

2.Experimental

2.1.Materials

The ceramic membrane,which was fabricated by coating anα-Al2O3membrane layer on the exterior surface of a tubular α-Al2O3support,was supplied by Jiangsu Jiuwu High-tech Co.,Ltd.,China.Hexadecyltrimethoxysilane(HDTMS)(≥85.0 vol%,Sigma-Aldrich Co.,LLC.USA)and ethanol(≥99.7 vol%,Wuxi City Yasheng Chemical Co.,Ltd.,China)were used for surface modification.HNO3(analytical grade)was obtained from Sigma-Aldrich Co.,LLC.(USA).NaOH(≥96.0 wt%)was provided by Xilong Chemical Co.,Ltd.,China.All chemicals were used without further purification.

2.2.Membrane modification

The procedure forthe hydrophobic modification oftubular asymmetric Al2O3membraneswas similarto ourprevious work given by Gao etal.[33]The Al2O3ceramic membrane tubes were soaked in deionized water for 2 h and dried in an electrically heated drying oven at 110°C for 6 h.After being cooled in air,the dried ceramic membrane tubes were immersed into 0.01 mol·L-1HDTMS ethanolsolution atroom temperature for 12 h.After the immersion,the Al2O3membrane tubes were rinsed with ethanoland deionized water to getrid ofunreacted HDTMS residue.The Al2O3membrane tubes were subsequently dried at 110°C for 2 h.

2.3.Membrane characterization

The membrane properties of the unmodified and modified Al2O3tubular ceramic membranes were characterized by scanning electron microscopy(SEM),contact angle measurement and gas permeation test.The crystal structure and phase composition of membranes were measured by X-ray diffraction(XRD).

The surface and cross-sectional morphologies of the modified and unmodified membranes were observed by scanning electron microscopy(SEM,S4800,Hitachi,Japan).The tubular membrane was cut off carefully to make a smooth cross-sectional surface.The cutting-off tubes were placed on a disc for sputtering with a thin filmof gold before testing.The SEM micrographs of the cross-section and the outer surface were taken at various magnifications.

The contact angles of water drops on the surface of the modified and unmodified membranes were measured using a DropMeter A-100(MAIST Vision Inspection&Measurement Co.,Ltd.,PRC)to quantify the wettability of the membrane surface.A 2 μl droplet of distilled water was dropped carefully onto the membrane surface,and the images of the droplet were recorded by a video camera at a rate of 100 frames per second.The contact angles were calculated from these images with the software named “Drop-meter”.The static contact angle of water on the modified Al2O3membrane surface,was also recorded in sessile drop mode.Each presented value is an average ofatleastthree measurements.

Gas permeation of the unmodified and modified membranes was measured.Nitrogen was used as the test gas for the gas permeation measurement.The ceramic membrane tube was covered by a re fined cylinder.Nitrogen(N2)was fed into the cylinderatdifferent pressures,and the gas permeation through the tube was measured by a soap bubble flow meter.

The crystalstructure and phase composition ofmembranesbefore and after long-term test were measured by X-ray diffraction(XRD,Smart Lab,Rigaku,Japan).The XRD patterns,scanned with 2θvalues from20°to 80°,were obtained using a step size of 0.02°under the condition of Cu Kαradiation(λ=0.154 nm)source at 40 kV and 40 mA.

2.4.SO2 absorption test

Fig.1.a.Schematic diagram ofSO2 absorption using a hydrophilic ceramic membrane contactor.b.Schematic diagram ofSO2 absorption using a hydrophobic ceramic membrane contactor.

Fig.1(a)and(b)respectively shows the schematic diagram of SO2absorption using a hydrophilic ceramic membrane contactor and a hydrophobic ceramic membrane contactor.Feed gas was prepared by diluting 2000 μl·L-1SO2ofSO2/N2mixture with pure N2.Gas flow rate was controlled by mass flow controllers(MFC D07-19B,Beijing Sevenstar Electronics Co.,Ltd.,PRC).The feed gas mixture, flowed downward,was introduced into the shell side of the membrane module.The absorbent,flowed upward,was fed into the tube side of the membrane module.The difference between Fig.1(a)and(b)was the position ofthe peristaltic pump.In Fig.1(a),the absorbent was sucked into the tube side of the membrane module with liquid pressure slightly lower than atmospheric pressure to avoid the absorbent penetrating through the pore of the membrane and flowing down along the wall of the membrane.In Fig.1(b),the absorbentwas forced into the tube side ofthe membrane module with liquid pressure slightly higher than atmospheric pressure to avoid the gas appearing on the liquid phase side.The inlet and outlet SO2concentrations were analyzed with a SO2gas analyzer(M60,Afriso,Germany).The specifications of the membrane contactor are listed in Table 1 and the details of tubular asymmetric α-Al2O3membranes and membrane module are shown in Fig.2.

Table 1 Specifications of the membrane contactor

Fig.2.Details of the membrane contactor.

In this study,the SO2removal efficiency and SO2mass transfer flux were used to evaluate the separation properties of membranes,which can be calculated as follows[34]:

where η is the SO2removal efficiency,%;J is the SO2mass transfer flux,mol·m-2·h-1;Qinand Qoutrepresent the inlet and outlet feed gas flow rates,respectively,m3·h-1;Cinand Coutare the SO2volumetric fraction in the gas inletand outlet,respectively;Tgis the gas temperature,K;and A represents the effective area of the membrane determined as follows[30]:

where L,dout,and dinare the effective length,outer diameter and inner diameter,respectively,of the Al2O3tube membrane utilized in these experiments,m.

3.Results and Discussion

3.1.Membrane characterization

The impact of hydrophobic modification on the surface and the cross-section morphology of the2O3membrane has been investigated.Morphologic features observed by SEM can be summarized as follows:solid-solid interfaces are clearly visible for both the original and the modified Al2O3membrane,and the thicknesses of the two membrane layers are approximately 50 μm.From Fig.3(a)and(c),there was no clear indication of the presence of a HDTMS film.The HDTMS molecules were only bonded to the surface but did not polymerize to form a thick layer to be seen because the number of hydroxyl groups on the surface of sintered Al2O3membrane was limited.As a result,the overall structures and the surface porosity of these two membranes hardly changed.This might be the reason why the gas permeance,as shown in Fig.5,slightly decreased after the HDTMS grafting.

As shown in Fig.4,the water contact angle of the original Al2O3membrane decreased from 54.9°to 0.1°within 3.5 s,indicating that the original Al2O3membrane was hydrophilic.In contrast,the water contact angle of the modified membrane can maintain around 130.6°,indicating that most of the hydroxyl groups on the membrane surface have reacted with HDTMS and the modified membrane was hydrophobic.

As shown in Fig.5,the results indicated that the modified Al2O3membrane tended to have slightly lower gas permeance because the HDTMS coating on the surface will increase the mass resistance for gas permeation through the membrane,which is consistent with the work of Koonaphapdeelert and Li[35].

3.2.Comparison between hydrophilic and hydrophobic Al2O3 membrane contactors

For the purpose ofevaluating the potentiality of hydrophilic ceramic membrane for SO2absorption,we compared the SO2absorption performance in the hydrophilic and hydrophobic Al2O3membrane contactors,respectively,at various conditions of NaOH concentration of 0 and 0.5 mol·L-1,inlet SO2concentration of 1000 μl·L-1,gas flow rate range of 120-1000 ml·min-1,liquid flow rate of 30 ml·min-1,and absorbent temperature of 20°C.As shown in Fig.6,with respect to the SO2-H2O system,the hydrophobic membrane contactor was found to be better than the hydrophilic one,while for the SO2-NaOH system,the result reverses.

Fig.3.SEMimages ofthe originalAl2O3 membrane and the modified Al2O3 membrane:(a)surface image oforiginalAl2O3 membrane;(b)cross-section image oforiginalAl2O3 membrane;(c)surface image of the modified Al2O3 membrane;(d)cross-section image of the modified Al2O3 membrane.

Fig.4.(a)Time dependence of water contact angles for the Al2O3 membrane and the modified membrane in atmosphere;(b)spreading and permeating behavior of water droplet.

Fig.5.The permeance of nitrogen through the membranes.

Fig.6.Comparison of absorption performance between the hydrophilic membrane contactor and the hydrophobic membrane contactor:(a)SO2 removal efficiency and(b)SO2 mass transfer flux(inlet SO2 concentration=1000 μl·L-1,Q L=30 ml·min-1,T L=20 °C).

These observations can be explained as follows.For the hydrophobic membrane,the SO2mass transfer occurred in three steps,i.e.,(i)diffusion from the bulk gas phase to the outer surface of the membrane,(ii)diffusion through non-wetted membrane pores,and(iii)dissolution into the absorbent.As for the hydrophilic membrane,the SO2mass transfer steps consisted of(i)diffusion in the gas phase that is the same as in the hydrophobic membrane contactor,(ii)dissolution into the absorbent on the outside surface of the membrane,and(iii)diffusion through wetted membrane pores to the bulk liquid phase.With respect to the SO2-NaOH system,the reaction of NaOH with SO2is considered to be instantaneous and irreversible if excessive free NaOH molecules exist.In the SO2absorption into NaOH in the hydrophilic membrane,the liquid in the pores contained so excessive NaOH molecules to absorb extremely low concentration of SO2,which ensured a timely renewal of free NaOH molecules from the bulk liquid to the gas-liquid interface.Consequently,the mass transfer of SO2in the membrane phase of the hydrophilic membrane was faster than that in the hydrophobic membrane.But for the SO2-H2O system,the SO2absorption into water is so slow that the reaction zone would extend beyond the liquid film and even to bulk liquid.In this case,the mass transfer of SO2in the hydrophilic membrane with liquid- filled pores was slower than that in the hydrophobic membrane with gasfilled pores.

To further validate these viewpoints,we performed experiments using solvents with various NaOH concentrations.As shown in Fig.7,increase of NaOH concentration can provide more hydroxyl(OH-)to absorb SO2,which is of great bene fit to enhance SO2removal efficiency.For SO2absorption in the hydrophobic membrane,the SO2removal efficiency did not increase any more when the NaOH concentration reached 0.1 mol·L-1.In contrast,for SO2absorption in the hydrophilic membrane,the SO2removal efficiency increased with NaOH concentration.Moreover,the hydrophilic membrane performed better than the hydrophobic one when the NaOH concentration was higher than 0.2 mol·L-1,concurrently showed a maximum level of SO2removal efficiency that was mostly double that of the hydrophobic membrane.These results showed that hydrophilic ceramic membranes have strong potential for use in the membrane SO2absorption process.

Fig.7.Comparison of SO2 removal efficiency between hydrophilic membrane and hydrophobic membrane using different concentrations of NaOH solution(inlet SO2 concentration=1000 μl·L-1,Qg=600 ml·min-1,Q L=30 ml·min-1,T L=20 °C).

Fig.8.Effect of gas flow rate on SO2 absorption into 0.2 mol·L-1 NaOH solution in hydrophilic Al2O3 membrane contactor(inlet SO2 concentration=1000 μl·L-1,Q L=30 ml·min-1,T L=20 °C).

Fig.9.Effect of absorbent flow rate on SO2 absorption using 0.2 mol·L-1 NaOH solution:(a)SO2 removal efficiency;(b)SO2 mass transfer flux(inlet SO2 concentration=1000 μl·L-1,T L=20 °C).

3.3.Effects of several parameters on SO2 absorption in hydrophilic Al2O3 membrane contactor

The effects of various operating variables including gas and liquid flow rates,absorbent temperature,inlet SO2concentration and NaOH concentration on the SO2absorption performance using hydrophilic ceramic membranes have been comprehensively investigated.

3.3.1.Effect of gas flow rate on SO2 absorption

The effects ofgas flowrate on the SO2removalefficiency and the SO2flux were investigated at an inlet SO2concentration of 1000 ppm,liquid flow rate of 30 ml·min-1,absorbent temperature of 20 °C,NaOH concentration of 0.2 mol·L-1and gas flow rate range of 120-1000 ml·min-1.Fig.8 shows that the SO2removal efficiency decreased slowly as the gas flow rate increases from 120 ml·min-1to 360 ml·min-1and then decreased quickly as the gas flow rate further increases.This is probably due to the fact that a low gas flow rate allowed the SO2to stay in the membrane contactor for a long time,which is favorable for SO2absorption.In contrast,a high gas flow rate led to a short residence time for SO2in the membrane contactor.The SO2mass transfer flux first increased linearly as the gas flow rate increased from 120 ml·min-1to 360 ml·min-1.This can be reasoned from the information that an increase in gas flow rate would decrease the boundary layer thickness of the gas phase in the shell side,resulting in increasing mass transfer performance.However,the increase of SO2flux slowed down as the gas flow rate further increased.This can be attributed to the finding thatthe proportion ofgas phase mass transfer resistance gradually dwindled to a small extent.

3.3.2.Effect of absorbent flow rate on SO2 absorption

Fig.9 shows the effects of absorbent flow rate,ranged from 30 ml·min-1to 180 ml·min-1,on the SO2removal efficiency and the SO2flux.Generally,increasing the liquid flow rate in MGA process may:(i)provide more free active absorbent molecules that enlarge the SO2absorption capacity and(ii)intensify the turbulent extent,which will greatly decrease the thickness of the liquid film,allowing the liquid-side mass-transfer coefficient to increase.It can be seen in Fig.8 that,however,the SO2mass transfer flux was almost equal,indicating that an increase in liquid flow rate from 30 to 180 ml·min-1had little effect on the SO2absorption performance.Combining with the observation that an increase of gas flow rate from 600 ml·L-1to 1000 ml·L-1(67%increase)enhanced the SO2mass transfer flux less than 5%,it can be inferred that the SO2absorption in a ceramic membrane is a membrane-phase controlled mass transfer process.

3.3.3.Effect of absorbent temperature on SO2 absorption

Absorbent temperature plays an important role in the mass transfer behavior.Fig.10 shows the effect of absorbent temperature on the SO2absorption.An increase in the absorbent temperature from 10°C to 35°C resulted in the SO2removal efficiency and the SO2mass transfer flux increasing from 48.6%to 88.2%and from 0.094 mol·m-2·h-1to 0.156 mol·m-2·h-1,respectively.This is because increasing absorbent temperature would enhance molecular diffusion and chemical kinetic,leading to promoting SO2absorption.

Fig.10.Effect of absorbent temperature on SO2 absorption using 0.2 mol·L-1 NaOH solution(inlet SO2 concentration=1000 μl·L-1,Q g=600 ml·min-1,Q L=30 ml·min-1).

Fig.11.Effect of inlet SO2 concentration on SO2 absorption:(a)SO2 removal efficiency;(b)SO2 mass transfer flux(Q g=500 ml·min-1,Q L=30 ml·min-1,T L=20 °C).

3.3.4.Effect of inlet SO2 concentration on SO2 absorption

In the presentwork,the inletSO2concentration was varied from400 to 2000 μl·L-1.As shown in Fig.11,it was found that the SO2removal efficiency became lower as the inlet SO2concentration increased.It can be reasoned that the higher the inlet SO2concentration,the more SO2was to be absorbed and the more reactive OH-was consumed.This also explained that(i)for the absorbent with 0.4 mol·L-1NaOH,the SO2removal efficiency started to significantly reduce until the inlet SO2concentration was over 1200 μl·L-1and(ii)for that with 0.2 mol·L-1NaOH,the SO2removal efficiency dramatically reduced as the inlet SO2concentration slightly increased.The SO2mass transfer flux enhanced as the inlet SO2concentration increased due to the increase of gas-phase driving force.However,the SO2mass transfer flux for those with 0.2 mol·L-1and 0.4 mol·L-1NaOH tended to reach their respective maximums as the inlet SO2concentration further increased.

3.3.5.Effect of NaOH concentration on SO2 absorption

The effect of Na OH concentration on SO2absorption performance was illustrated in Fig.12.When the Na OH concentration was lower like 0.001 mol·L-1,the SO2absorption performance was similar to water as absorbent.It could be found that the SO2absorption performance was not sensitive to the lower concentration of absorption solution.However it became sensitive to the concentration when the NaOH concentration was more than 0.1 mol·L-1NaOH solution.A higher concentration of NaOH solution resulted in a better absorption performance.This is because the amount of OH-in the gas-liquid interface became more and the diffusion rate of OH-from the liquid bulk to the liquid-gas interface rose with increasing OH-concentration gradient.Thus,enhancing NaOH concentration is favorable to increasing the absorption rate and capability of solution.

Fig.12.Effect of NaOH concentration on SO2 absorption:(a)SO2 removal efficiency;(b)SO2 mass transfer flux(inlet SO2 concentration=1000 μl·L-1,Q L=30 ml·min-1,T L=20 °C).

3.4.Long-term stability of hydrophilic ceramic membrane for SO2 absorption

A good membrane contactor should be capable of long-term stability in mass transfer performance.SO2absorption performance over 480 h was performed via a hydrophilic Al2O3membrane contactor using 0.5 mol·L-1NaOH in a closed loop to evaluate its long-term performance.After 360 h,part of the recycled absorbent was drained and 1 mol·L-1NaOH solution was added to increase the concentration of NaOH.As shown in Fig.13,for the hydrophilic Al2O3membrane,the SO2removal efficiency and the SO2mass transfer flux maintained 100%and 0.151 mol·m-2·h-1,respectively,within 264 h.Then,the SO2absorption performance tended to decline and the SO2removal efficiency dropped to 90%on the operation time of 360 h due to the decrease of effective OH-concentration.With high concentration of NaOH solution adding into the absorbent,the SO2removal efficiency and the SO2mass transfer flux reverted to their original levels in the next 120 h.

Fig.13.Long-term stability of hydrophilic ceramic membrane for SO2 absorption using 0.5 mol·L-1 NaOH solution(inlet SO2 concentration=1000 μl·L-1,Q g=600 ml·min-1,Q L=30 ml·min-1,T L=20 °C).

The SEM images of membrane surfaces before and after longterm test were shown in Fig.14.There were some sodium hydroxide crystal on the surface of the tested membrane after washing with water,but no distinct microcracks or other great defects containing exfoliation of crystal grains can be observed in the microstructure of membranes.The XRD patterns of the membranes before and after long-term test were shown in Fig.15.It was obvious that the crystal phase of the membranes before and after long-term test was not changed,with peaks of α-phase sharp.The above results indicate that the α-Al2O3membranes had strong alkali resistance.It is reasonable to conclude that the hydrophilic α-Al2O3membrane shows exceptional long-term stability in the SO2absorption process.

Fig.15.The XRD patterns of the membranes before and after long-term test.

4.Conclusions

The performance of SO2absorption into NaOH solution was experimentally investigated in a hydrophilic Al2O3membrane,and compared with that in a hydrophobic Al2O3membrane.The results showed that the hydrophilic membrane was superior to the hydrophobic one when the NaOH concentration was higher than 0.2 mol·L-1.In addition,it was found that the mass transfer of SO2absorption in the hydrophilic membrane was controlled by gasfilm resistance when a high NaOH concentration was used,while by membrane- film resistance when using a low NaOH concentration.Moreover,the stability of the hydrophilic Al2O3membrane was confirmed by 480 h test of SO2absorption.Although hydrophilic ceramic membranes for gas absorption are certainly still in their initial state,they show potential for post-combustion SO2absorption.

Fig.14.SEM images of the membranes before and after long-term test:(a)before long-term test;(b)after long-term test.