A Comparative CO2 Adsorptionanalysis Inpure and Amine Modifiedcomposite Membranes

A comparative CO2 adsorptionanalysis inpure and amine modifiedcomposite membranes.

SarahFarrukh[1], 2*, X. Fan 2, Arshad Hussain 1

1. School of Chemical and Materials Engineering (SCME), National University of Sciences & Technology, 45000, H-12, Islamabad, Pakistan.

2.Institute for Materials and Processes, School of Engineering, University of Edinburgh, EH9 3JL, Scotland, UK.

Abstract: In this study, theCO2adsorption analysis inCA-TiO2andCA- APTMTiO2blended membranes was done. The membranes were also characterised using SEM and FTIR analysis techniques.The adsorptionresults indicated that 120˚C and 90 ˚C were considered as optimised temperature for regeneration of CA-TiO2 andCA- APTMTiO2 membranes. Thetesting results revealed that adsorption capacity reached maximum at 3.0bar.Validation of experimental results was done by Pseudo first order, Second order andIntra particle diffusion models. The correlation factor R2 represented that the second order model was fitted well with experimental data. The intra particle diffusion model represented that adsorption isnotsingle step process.

.

Keywords: cellulose acetate(CA); gas adsorption; pseudo order models; CO2; TiO2; composite membranes.

Introduction

The yearly increment in the concentration of CO2 in the atmosphere, enhancing global warming and resulted in drastic climate changes. The Antarctica’s ice core study revealed that since 650,000 years, the earth’s temperature and concentration of CO2 has been doubled. It is also proposed that the concentration of CO2 will be reached at 600 ppm by the year 2050. So it is extremely important to stabilize the CO2concentration in the range of 450-750 ppm by reducing the human activities or to capture it by utilizing various CO2 capture techniques such as membrane gas separation, cryogenic distillation, amine absorption, membrane gas absorption and adsorption processes[1]. The above mentioned techniques are being used in power generation, industrial and fuel(natural gas) processingareas to control the concentration of CO2. However, in pre and post combustion CO2 capture, membrane processes have been emerged as a fast growing gas separation technology due to low capital and operation cost, low energy requirement, environment friendly, small foot print and flexibility in handling of higher flow rate, pressure and various feed composition [2].

During last decade, CO2 adsorption technique, utilising particles or fillers as adsorbent, is under investigation.Various adsorbents have been synthesized to study adsorption of CO2 gas on them. The activated carbon has been extensively utilised due to low cost, thermal stability and less sensitivity towards moisture [3].In recent years, Zeolite[4], alumina [5] and silica[6]have produced remarkable results in CO2 adsorption.Lately, Metal –organic frameworks (MOFs) have attracted the attention of many researchers in recent years due to high CO2 capture efficiency. But the main drawback of low CO2 selectivity persuades the researchers to work with amine functionalised adsorbents[7].

In recent years, amines are considered extremely useful in capturing of CO2 because the acidic gas such as CO2 is considered to be more attracted towards amines as compared to other functional groups. This is the basic principle behind amine absorption CO2 capture. This principle was further utilized in membrane technology and in gas adsorption technology by incorporating amine functional groups in polymers, organic and in-organic fillers[8-9].In fact, the sorption capacity of CO2 by amines is the major factor, which affect the CO2 capture performance of membranes or adsorbents.

As it is already known that the separation of CO2 through dense membranes is dependent on solution-diffusion mechanism other than sieving out of gas via membranes. The dual mode sorption model is used to describe sorption and penetration of gas molecules in glassy polymeric membranes. [10-11].Thismodel comprises of a Henry's law ‘dissolved’ mechanism and Langmuir ‘hole-filling’ mechanism. Henry’s law scrutinised that the sorption of gas in polymer matrix, depends upon the dissolution of gas in the polymer matrix, whereas Langmuir model demonstrate about the sorption of gas molecules in voids or molecular scale cavities of polymer matrix[12]. The above mentioned laws are important in a respect that in glassy polymeric membranes, gases sorbed, dissolved and adsorbed. This adsorption can affect the permeability and selectivity of gases through membranes.

The gas transport models in glassy polymer encourage us to study the CO2 gas adsorption in dense glassy polymericcomposite membranes to check either CO2 gas adsorbed in these membranes or not. In our recent work, CO2 adsorption study was done for pure Cellulose acetate membrane andcellulose acetate-titania(CA-TiO2)composite membrane. The results were surprising as CO2 gas was adsorbed more in CA-TiO2 composite membrane as compared to pure CA membrane[13].However, the sorption phenomenon in amine modified composite membranes is still under investigation.

Considering these factors, itwas planned to study CO2 adsorption in CA- APTM (3-aminopropyl-trimethoxysilane) modified TiO2composite polymeric membranes and then to compare the results with CA-TiO2 composite membranes. The cellulose acetate (CA) was utilised, because it is one of the glassy polymers that are readily available, economical, strong and easily handled[14]. TiO2nano particles were considered because they have been observed as one of the adsorbents that can enhance CO2adsorption[15-17]. APTM was selected to modify TiO2nano particles by incorporating NH2 groups in them. It was intended to check either this modification can affect CO2 adsorption in membrane as compared to CA-TiO2 membrane or not.The membranes were fabricated using the Diffusion Induced Phase Separation (DIPS) described already[18]. The CA was blended with 20 wt% of pure and APTM modifiedTiO2 nanoparticles. The morphological studies were done by SEM, whereas qualitative studies of membranes were done by FTIR analysis. The data was modelled using pseudo first order, pseudo second order and inter particle diffusion models.

1. Material and methods

1.1. Chemicals

Cellulose acetate (CA) (Sigma Aldrich) was the membrane forming material. Acetone (BDH Laboratory) was used as solvent for CA. Isopropyl Alcohol (Riedel- de Haen, Sigma Aldrich) was used for drying membranes. TiO2nano particles (25nm) were bought from Sigma Aldrich, UK where as APTM (Sigma Aldrich UK) was used as NH2 cross linker.

1.2. Modification of TiO2nano particles

The modification of TiO2nano particles was done according to the reported procedure [19]. TiO2 particles and APTM (3-aminopropyltrimethoxysilane) were ultrasonicated for 30 min. The mixture was degassed using nitrogen gas. The mixture was magnetically stirred at room temperature for 6 hrs. The product powdered was separated by centrifugation and dried in vacuum at 80˚C for 24 hrs.

1.3. Fabrication of pure and modified composite Membranes

The composite membranes were synthesized by dissolving 1.5 g of CA in 3/4th of the solvent (6.7 g) and kept on stirring. 0.3g of pure and amine modified TiO2nano particles were ultrasonicated in remaining solvent (2.15g) for 2 hours and added to the CA mixture along with stirring[13]. The blended mixture was casted on glass slab after the removal of bubbles in mixture. The gelling process of membranes was done by dipping membrane is cold water for 20 min. In next step membrane was dipped in hot water 80˚C for 30 minutes. Isopropyl alcohol was utilised to dry the membranes. The CA- TiO2and CA- APTM TiO2blended membranes were considered in this study, which labelled as M(a) and M(b) respectively.

1.4. Physical Characterization

The surface morphologies of membranes were investigated by Scanning Electron Microscopy (SEM, S-4700 Hitachi, Japan). The fabricated membranes were attached with double sided tape on the surface of Aluminium stubs. In next step, the upper surfaces of membranes were sputter coated with thin gold film for one minute and then placed inside the SEM chamber to study the morphologies of membranes.

Fourier Transform Infrared Spectroscopy (FTIR) was utilised for qualitative structural analysis of membranes, with 1 cm-1 resolution in transmission mode. The wave number varies from 450 to 4000cm-1.The synthesized membranes were cut in small pieces and placed in pallet holder. The holder was mounted in FTIR instrument and all the spectra subtracted from background were recorded at room temperature.

1.5. Adsorption setup

The adsorption studies of fabricated membranes were performed by using locally made gas adsorption rig as represented in Figure1. According to the Fig1, the adsorption cell was connected with gas cylinders (CO2 and N2) through stainless steel pipes. The entire setup was kept in temperature controlled environment with proper exhaust ventilation (Nederman).The gases which utilized in this study, were 99.9% pure and regulated by Lab Master Pressure regulator. To control the required gas flows and pressures,the flow meters and pressure gauges (Scotia Instrumentation) were connected above and below the adsorption cell. The CO2gas sensor, belong to Cozir series, was also connected at the other end, which was further connected with computer to read the adsorption data. The diameter of adsorption cell was 2 cm and the apparatus was tested for leak absence [13].

1.6. Experimental Procedure

Before gas adsorption analysis of membrane, the whole setup was cleaned up by purging it with N2 gas for two hours. The flow rate of N2 gas was kept 100cc/min under the pressure of 1 bar at 25 ˚C

The membranes were cut in circular shape and diameter was kept 2 cm. The mass of these membranes measured using analytical balance as a reference. The membranes were given the heat treatment by heating them at 120˚C and 90˚C to avoid any contamination and water vapours. Both the temperatures were optimised for two different membranes.The mass of membranes with and without membrane module was measured to know the mass of pure membranes.

After weight measurement of membranes and purging of adsorption setup, membranes were placed on cell and adsorption study of gas was done at different pressures such as 1.0, 1.5, 2.0, 2.5, 3.0, 3.5 and 4.0 bars at 25˚C with constant flow rate 60 cc/min. During CO2 gas passage through membrane cell, the CO2 sensor was utilised to measure the concentration of CO2.

The amount of CO2 adsorbed on the membrane was measure using equation(a)

(a)

Where q is amount of CO2 adsorbed (mg/g), mois the original mass of adsorbent andmt is the mass of adsorbent, at time t[13].

2. Results and discussions

2.1. SEM analysis

To study the surface and cross sectional morphologies of blended membranes, SEM images were captured and representedin Fig. 2(a)and2(b). The surfaces of membranes were studied at 5000* magnification, whereas cross sectional images were taken at 10,000 *magnifications. According to Fig.2(a), the surfaces of CA-TiO2and CA-APTMTiO2are dense but not completelysmooth. This is due to the fact thatblended membranes have homogenously dispersed TiO2nano particlesbut with some nano scale granular appearance as stated by Sijbesma et al., 2008[20].In the case of CA- APTM TiO2 blended membrane, the surface of membrane has some valleys other than granular appearance. By viewing at high magnifications, it can be stated that these are depressions but not pores.

In Fig.2(b), the cross sectional images of synthesizedmembranes are also depicting the dense structure of membranes.It is well recognized that the TiO2nano particles are hydrophilic in nature and during solvent –water exchange step, they can enhance the thermodynamic instability in membrane, which can produce porosity in membranes, therefore the concentration of TiO2nanoparticles was optimised to avoid the generation of porosity in membranes.

Insert Figure 2(a) and 2(b)

2.2. FTIR analysis

The qualitative structural analysis of membranes was performed by Fourier Transform Infrared spectroscopy to observe the functional groups in synthesized membranes. The FTIR spectra of CA-TiO2 and CA- APTMTiO2 are depicting in Fig.3. In the spectrum of both CA-TiO2and CA-APTM TIO2 membranes, the vibration band in the range of 3550-3100 cm-1 corresponds to –OH stretching. However, the characteristic vibration bands of C=O and C-CH3 are observed in the range of 1790-1720 cm-1 and 1370 cm-1 respectively. Moreover, the single bond of ether C-O-C stretching vibration is related to the band at 1235 cm-1. These peaks are confirming the presence of Cellulose Acetate polymer in the membrane.

Conversely, the distinctive peak centred at 800-500 cm-1is due to the vibration of Ti-O bond, confirming the presence of dispersed TiO2nano particles in CA matrix[21]. In the spectrum of CA- APTM TiO2membrane, the peaks at 800-500 cm-1 are confirming the presenceTi-O bonds; however this peak is broader than the peak of CA-TiO2 membrane. This probably is due to interaction of amine with Ti-O bonds. As according to Stefaniaetal[22], the peak at 850–750 cm−1corresponds to wagging and twisting NH2 bonds and peak at 715 cm-1represents wagging N–H group. Another peak in the range 1550-1600 cm-1 is depicting the presence of asymmetric and symmetric bending of primary amine(NH2). These peaks indicate the incorporation ofaminegroup on the surface of TiO2nano particles which were further blended with CA polymer. This spectrum is supported by Sakpaletal. 2012[23].

InsertFigure 3

2.3. Regeneration of membranes

According to Kaithwasetal., 2012,[24]the temperature range of flue gases in post combustion processes is in the range of 50˚C to 120˚C. For that reason, it is extremely important to study the behaviour of membrane under different temperatures at different times. In this section, experiments were designed to investigate the effect of temperature on membranes. The fabricated membranes were heated at different temperatures to get the complete picture of heating effects in membranes.

The Fig.4(a) and 4(b) are representing the % mass loss in membranes at different temperatures.In general, the heating resulted in a mass loss of the membranes, but the profile varies with the temperature. For both synthesized membranes, the loss of mass reaches maximum after 10 s of heat treatment for all temperatures, 700C to 1200C. According to Fig.4(a) and 4(b), it can be observed that for CA-TiO2 membrane, the % mass loss at 120˚C is in normal range, it is reflecting the removal of water, CO2 and other impurities. Whereas, the same mass loss can be observed at 90 ˚C for CA-APTM TiO2 membrane. However, at high temperatures such as 110 ˚C and 120˚C, the maximum mass loss is observed in CA-APTM TiO2 membrane. This probably is due to loss of amine at these temperatures. As stated by Gebaldetal. 2011[25], that amine based adsorbents are suitable for the adsorption of CO2, but this functional group can be released by heating at 100 ˚C. Moreover, in industrial processes the CO2 is removed by heating adsorbent at 90˚C ±1˚C. Due to these reasons the optimised temperature for CA-APTM TiO2 membrane is considered 90˚C. While, 120˚C is optimised temperature for CA-TiO2 blended membrane.

Insert Figure 4(a)and 4(b)

After the selection of optimised temperatures for the regeneration of fabricated membranes, it was extremely imperative to explore the time required for maximum mass loss of membranes. So to spot the time of maximum mass loss in membrane at optimised temperatures(120 ˚C for CA-TiO2 and 90˚C for CA-APTM TiO2), differential curves (mass loss with respect to time) were calculated for both of membranes given in Fig.5(a) and5(b) .The differential curves are depicting important result that for CA-TiO2 membrane the maximum mass loss was done within 5 s at 120 ˚C, whereas for CA-APTM TiO2 at 90 ˚Cthe maximum mass loss was occurred within 10 s of thermal treatment. This probably represents that the maximum removal of solvent, water vapours, humidity and CO2 from the membrane was done in first 5 to 10 seconds of heat treatment of membrane, where as it took further 50 seconds to be in equilibrium.

Insert Figure 5(a) and 5(b)

2.4. Adsorption Capacity

The CO2 gas adsorption and its kinetics were investigated because slower process determines the overall uptake rate[26]. The adsorption isotherms of blended membranes at seven different pressures are given in Fig.6 (a). According to these Isotherms, both blended dense membranes have adsorbed CO2 gas. Interestingly, it can be observed from the curves that even at low pressures, the CO2 adsorption in CA- APTM TiO2 blendedmembrane is high as compared to CA-TiO2 blended membrane.The CO2 adsorption trend is increasing linearly for both membranes, till it reaches maximum at 3 bars. From 3bars the adsorption of CO2becomeconstant, but still high for amine modified composite membrane (CA-APTM TiO2).

Insert Figure 6(a)

The CO2adsorption inCA–TiO2 blended membranewasprobably due the acetate and ester groups of CA which enhance the CO2 affinity, as explained byDonohue (1989).Secondly,the addition of TiO2nano particles has increased the free volume between polymeric chains. This free volume may be helpful in diffusion and adsorption of CO2 in polymeric matrix. Besides free volume and acetate group of CA, the OH functional groups also have loving affinity for CO2 as compared to other gases and TiO2 has OH functionality to adsorb CO2. Another significant aspect is polarity, it is also important for gas and polymer interaction. As CO2 has quadrupole moment so it is soluble in polar polymers and its polar bonds have strong interaction with OH functional groups of TiO2. These experimental findings are also supported by previously published results in which TiO2 was blended with PI, PVA, PAI[27- 29].

However, still this adsorption capacity of CA-TiO2 membrane is low as compared to CA- APTMTiO2 blended membrane. In CA- APTMTiO2 blended membranes, other than the collective effects of free volume in polymer matrix due to TiO2nano particles; acetate group of CA; OH functionality and polar bonding, there is amine functionality. This functionality has strong affinity for CO2 gas molecules as supported by the work of many researchers. So with this N-H functional group, CO2 molecules were attracting more towards CA- APTMTiO2 membrane to get adsorbed.

2.5 Adsorption kinetics

The transient CO2 adsorption profiles of CA-TiO2 and CA-APTM TiO2blended membranes are presented in Fig.6(b). According to this Figure, it can be observed that CO2 gas has been adsorbed by both fabricated membranes, but interesting results were achieved when CA-APTM TiO2 membrane has adsorbed more CO2 then CA-TiO2 membrane. This reflects that beside OH functionality and free volume, the amine functionhas increased the uptake of CO2in membrane. The CO2 adsorption was very intensive during the first 60 seconds in both membranes and after this it started to be constant. There is possibility that,the gas molecules make contact with surface in the first 10 seconds, then diffuse in membrane to get adsorb with in it, as stated by the Li et al., 2008 that the external surface of adsorbent cause rapid adsorption in start, after this slow diffusion and solubility process take place[30].

Insert Figure 6 (b)

To attain the adsorption uptake rate and residence time for the complete adsorption process, it is significant to explore the adsorption kinetics. This analysis is based on pseudo first order and pseudo second order reaction mechanisms[27,31]. The equation (b) is the pseudo first order equation.

(b)

Whereas, the pseudo second order equation can be written as