Adsorption

of

Isopropyl

Alcohol

by

Carbon

Nanotubes Paper # 1133 Shihchieh Hsu, Mengshan Lee and Chungsying Lu Department of Environmental Engineering, National Chung Hsing University, Taichung 40227, Taiwan

ABSTRACT Carbon nanotubes known as a novel discovered form of carbon. They are also expected to be excellent adsorbents for adsorption of trace pollutants from water or air. The objective of this paper is to study adsorption characteristics of isopropyl alcohol (IPA) vapor in the range of 40-800 ppmv from air stream with single-walled carbon nanotubes (SWCNTs). The test results indicated that the nature of the SWCNTs surface was greatly improved after purification by concentrated hypochlorous acid, which made SWCNTs become more hydrophilic and suitable for adsorption of IPA vapor from air stream. A significantly higher adsorption capacity of IPA vapor was found with purified SWCNTs than with raw SWCNTs. The BET isotherm is the best model to describe adsorption behavior of IPA vapor onto purified SWCNTs, followed by the Freundlich isotherm, and then the Langmuir isotherm. Keyword: Single-walled carbon nanotubes, adsorption, isopropyl alcohol.

INTRODUCTION Volatile organic compounds (VOCs) are among the most common pollutants emitted by chemical process industries dealing with manufacture and process of chemicals.1 Isopropyl alcohol (IPA), kind of VOCs, a widely used solvent and drying agent was chosen as a representative compound for low molecular weight, polar organic molecules.2 With increasing environmental awareness and stringent regulations of pollutants emission, adsorption offers an effective means to control VOCs emissions in low concentration levels.3 Since carbon nanotibes (CNTs) were first discovered by Ijima,4 the interaction of CNTs with their environmental, especially with gases or dopants adsorbed on their interior or exterior surface, has attracted increasing attention due to the anticipated influence on key properties of these materials.5 For the potential application of CNTs

in environmental field, they are relatively new adsorbents for adsorption of trace pollutant from water or air. Long and Yang6 reported that a significantly higher dioxin removal efficiency is found with CNTs than with activated carbon. CNTs have been proved to possess great potential application in environmental protection by foregoing investigator. However, the studies on the adsorption of trace pollutants with CNTs are still limited in the literature. The aim of this paper was to study adsorption characteristics of IPA vapor with single-walled nanotubes (SWCNTs) in the concentration range of 40-800 ppmv.

MATERIAL AND METHOD Preparation of purified SWCNTs Single-walled carbon nanotubes (L-SWCNTs, Nanotech Port Co., Shenzhen, China), with the diameter less than 2 nm and the length range 5-15 µm, were selected as adsorbents in this study. The mass ratio of amorphous carbon was less than 5%. The raw SWCNTs were dispersed into a 150 ml flask containing 70% concentrated hypochlorous acid solutions. The volume ratio of acid solution to SWCNTs was kept at 40. The mixed solution was refluxed using an ultrasonic cleaning bath (Model D400H, Delta Instruments Co., USA) at 85oC for 3 h to remove metal catalysts and amorphous carbon. After cooling, the SWCNTs were washed by deionized water. Finally, the SWCNTs-containing solution was filtered by 0.45 µm glass-fiber filter to obtain purified SWCNTs.

Experimental set-up The experimental set-up for adsorption of IPA by SWCNTs is shown in Fig. 1. The adsorption column was made of Pyrex and had a length of 30 cm and an internal diameter of 1.5 cm. A 3-cm headspace was designed for the IPA inlet while a 3-cm bottom space was designed for the outlet of treated air. The adsorption column was filled with 0.7 g SWCNTs (packing height=2.3 cm) and hold by a glass supporter in the middle of the column. The column was placed within a temperature control box (Model CH-502, Chin Hsin, Taipei, Taiwan) to maintain gas temperature at 25oC.

Figure 1 Schematic diagram of experimental setup

Compressed air was passed first through a filtration device (LODE STAT compressed air dryer, Model LD-05A, Taipei, Taiwan) to remove moisture, oil and particulate matter. After purification, the minor air stream was passed through the first glass bottle containing pure IPA solution (J. T. Baker, NJ, USA, 99.9% purity) to produce IPA vapor. The IPA vapor was then mixed with the major air stream in the second glass bottle and was passed downwards into the adsorption column. The influent IPA concentration was controlled by regulating the minor air stream rate using mass flow controller, while the empty-bed residence time (EBRT) was controlled by regulating the major air stream rate using mass flow controller (Model 247C four channel read-out and mass flow controllers, MKS instrument Inc., MA, USA). The variations in the influent IPA concentration were within 10% and the air stream was controlled at 0.217 lpm (EBRT=1.12 s). The effluent air stream was then flowed into an auto sampling system with a gas chromatograph (GC, Model SRI 8610C gas

chromatograph, SRI Instruments, CA, USA) equipped with a flame ionization detector (FID). Isopropyl alcohol adsorbed by SWCNTs was calculated as follows:

q=

1 t Q ⋅ (C in − C eff m ∫0

) dt

(1)

where q is the amount of IPA adsorbed by SWCNTs (mg/g); Q is the influent flow rate (lpm); Cin is the influent IPA concentration (mg/l); Ceff is the effluent IPA concentration (mg/l); m is the SWCNTs dosage (g) and t is the contact time (min).

Analytical methods Concentration of isopropyl alcohol was determined using a GC-FID. A 15 m fused silica capillary column with 0.32 mm inner diameter and 1.0 µm film thickness (Supelco wax, Supelco Inc., PA, USA) was used for isopropyl alcohol analysis. The GC-FID was operated at injection temperature of 150 oC, detector temperature of 200 o

C and oven temperature of 130 oC.

The total surface area, mean pore size, pore size distribution and pore volume of purified SWCNTs were determined by a BET sorptometer (Model BET-202A, Porous Materials Inc., NY, USA).

RESULTS AND DISCUSSION Characterization of purified SWCNTs Shape and size are essential as identifying characterization of adsorbents since these parameters determine the specific surface area and thus influence adsorption rate and sensitivity to environmental conditions. Figure 2 shows the scanning electron microscope (SEM) images of raw and purified SWCNTs, respectively. As can be seen, the cylindrical-shapes of isolated SWCNTs were not changed after purification process. However, it is evident that there are many metal catalysts attached on the surface of raw SWCNTs. Due to inter-molecular force, the isolated SWCNTs of different size and direction form an aggregated structure. Figure 3 shows the high resolution transmission electron microscope (HRTEM) images of raw and purified SWCNTs. As can be seen, the aggregated structure of raw SWCNTs was destroyed after purification process.

Figure 2 SEM images of CNTs: (a) raw SWCNTs, (b) purified SWCNTs.

Figure 3 HR-TEM images of CNTs: (a) raw SWCNTs, (b) purified SWCNTs.

Figure 4a and 4b exhibit the pore size distribution of raw and purified SWCNTs, respectively. It is obvious that the pore size of raw SWCNTs is a bimodal distribution. The major peak is located in the size range 4-50 nm while the minor peak is located in the size range 1-3 nm. The 1-3 nm pores are the SWCNTs inner cavities and responsible for around 25% of the total pore volume. The 4-50 nm pores are likely to be contributed by aggregated pores which are formed by the confined space among the isolated SWCNTs 7 and responsible for around 74% of the total pore volume. After purification, the pore radius in 1-3 nm range increased from 25 to 67% of the total pore volume while the pore radius in 4-50 nm range decreased from 74 to 30% of the total pore volume. This is due to the isolated SWCNTs were separated apart after purification process. Particularly, the pore volume for the pore radius less than 2 nm significantly increased after purification process, likely because of the removal of metal catalysts and amorphous carbon. Table 1 lists the BET measurement results of raw and purified SWCNTs. As can be seen, the specific surface area, the total pore volume and the average pore radius decreased after purification process. Table 1. BET measurement results of raw and purified SWCNTs

Specific surface area m2/g

Total pore volume cm3/g

Average pore radius nm

Raw SWCNTs

590.5418

1.1231

3.8037

Purified SWCNTs

423.1692

0.50423

2.0613

Figure 5 shows the Fourier transform infrared (FTIR) spectroscopy of raw and purified SWCNTs. It is seen that the raw SWCNTs exhibit less significant peaks. In contrast, the purified SWCNTs have more evident peaks at wavenumber of 1600, 1750 and 2850-2910 cm-1 which are associated with aromatic ring groups (quinone), carboxylic acids and phenolic groups (O-H), asymmetric and symmetric C-H stretching vibration in aliphatic, respectively.8, 9 It is evident that there are many oxygen-containing groups attached on the surface of purified SWCNTs.

Figure 4a Pore size distribution of raw SWCNTs

Volume Fraction (cm3/g)

2.0

1.5

1.0

0.5

0.0 1

10

100

Proe Radius (nm)

Figure 4b Pore size distribution of purified SWCNTs

Volume Fraction(cm3/g)

2.0

1.5

1.0

0.5

0.0 1

10

Pore Radius (nm)

100

Figure 5 FTIR spectra of CNTs: (a) raw SWCNTs, (b) purified SWCNTs.

46 44 42

b

quinone

Transmission(%)

40 38 36 34

-COOH

32 30

C-H

a

28 26 24 22 4000

3500

3000

2500

2000

1500

1000

500

-1

Wavenumber (cm )

Effects of contact time Figure 6a and 6b show the effect of contact time on the adsorption of 40 and 800 ppmv IPA vapor onto raw and purified SWCNTs, respectively. It is noted that the adsorption of IPA vapor increases quickly with time and then reaches equilibrium. The contact time for adsorption IPA onto raw SWCNTs to reach equilibrium is equal to 480 min for C0=40 ppmv and equal to 120 min for C0=800 ppmv. The final adsorption capacity of IPA reaches 16.21 mg/g for 40 ppmv and achieves 53.33 mg/g for C0=800 ppmv. The contact time for adsorption IPA onto purified SWCNTs to reach equilibrium exceeds 1020 min for C0=40 ppmv and equal to 150 min for C0=800 ppmv. The final adsorption capacity of IPA reaches 43.12 mg/g for 40 ppmv and achieves 79.36 mg/g for C0=800 ppmv. The longer contact time to reach equilibrium for lower initial IPA concentration may be explained by the fact that diffusion mechanism control the adsorption of IPA onto SWCNTs. Reid et al.10 indicated tat the mass diffusivity decreases with decreasing concentration under very dilute solution and causes the decrease of diffusion flux of adsorbate onto the surface of the adsorbent. As compared to the adsorption of IPA with raw and purified SWCNTs, it is evident that it takes longer contact time for purified SWCNTs.

Furthermore, the adsorption capacity of IPA onto purified SWCNTs is higher than onto raw SWCNTs. This can be attributed to more oxygen-containing groups attached on the purified SWCNTs surface, which make purified SWCNTs become more hydrophilic and suitable for adsorption of IPA vapor.

Figure 6a Effect of contact time on adsorption of IPA onto SWCNTs with the initial concentration (C0) of 40 ppmv 45 40

Capacity (mg/g)

35 30 25 20 15

Purified S WCNTs

10

Raw S WCNTs

5 0 0

200

400

600

800

1000

1200

Contact time (min)

Figure 6b Effect of contact time on adsorption of IPA onto SWCNTs with the initial concentration (C0) of 800ppmv 80 70

Capacity (mg/g)

60 50 40 30 Purifie d SWCNTs Raw SWCNTs

20 10 0 0

30

60

90 Contact time (min)

120

150

180

Adsorption isotherms Figure 7 shows the adsorption isotherms of IPA onto purified SWCNTs. The adsorbed amounts of IPA vapor are equal to 43, 54, 68, 74 and 79 mg/g, respectively, for the equilibrium concentrations of 40, 99, 291, 468 and 817 ppmv. The experimental data for IPA adsorption onto SWCNTs could be approximated by the isotherm models of Langmuir (2), Freundlich (3), BET(4) q=

q m k a Ce 1 + bC e

q = K f Ce

q=

(2)

n

(3)

qm k B Ce (1 − Ce / C s )(C s − Ce + k B Ce )

(4)

where q is the mass of IPA adsorbed by SWCNTs; Ce is the equilibrium IPA concentration, Cs is the saturation concentration; a and b are Langmuir constants; Kf and n are Freundlich constants; and kB is BET constants. The isotherm constants were obtained from fitting the adsorption equilibrium data and listed in Table 2. As can be seen, the BET isotherm is the best model (R2=0.9992) to describe the adsorption of IPA vapor onto SWCNTs, followed by the Freundlich isotherm (R2=0.9858), and then the Langmuir isotherm (R2=0.9706).

Figure 7 Adsorption isotherms for IPA with purified SWCNTs 90 80

Adsorption Capacity (mg/g)

70 60 50 40 30 20 10 0 0

200

400

600

Conccentration (ppmv)

800

1000

Table 2. Constants of Langmuir, Freundlich and BET isotherm models for adsorption of IPA onto purified SWCNTs Adsorption isotherm models qm Langmuir model ka R2 nF Freundlich model kF R2 qm BET model kB R2

Values 78.0858 0.0296 0.9706 0.2049 20.8113 0.9858 81.6622 913.3160 0.9992

CONCLUSIONS The following conclusions could be drawn from this study: 1.The inner pore volume of SWCNTs increased and the aggregated pore volume of SWCNTs decreased after purification by concentrated hypochlorous acid. 2.More oxygen-containing groups were found on the surface of purified SWCNTs, which made purified SWCNTs become more hydrophilic and suitable for adsorption of IPA vapor. 3.A significantly higher adsorption capacity of IPA vapor was found with purified SWCNTs than with raw SWCNTs. 4.The BET isotherm is the best model to describe adsorption behavior of IPA vapor onto SWCNTs, followed by the Freundlich isotherm, and then the Langmuir isotherm.

REFERENCES 1. Gupta, V.K.; Verma, N. Chemical Engineering Science. 2002, 57, 2679-2696. 2. Rana, N.; Raghu, P.; Shero, E.; Shadman, F. Applied Surface Science. 2003, 205, 160-175. 3. Dwivedi, P.; Gaur, V.; Sharma A.; Verma N. Separation Purification Technology. 4. 5. 6. 7.

2004, 39, 23-37 Iijima, S. Nature. 1991, 354, 6348, 8-56. Collins, P.G.; Bradley, K.; Ishigami, M.; Zettl A. Science. 2000, 287, 1801-1816. Long, R.Q.; Yang, R.T. J. Am. Chem. Soc. 2001, 123, 2058-2059. Yang, Q.H.; Hou, P.X., Bai, S.; Wang, M.Z., Cheng, H.M. Chem. Phys. Lett. 2001, 345, 18-24.

8. Ei-Hendawy, A.A. Carbon. 2003, 41, 713-722. 9. Pradhan, B.K.; Sandle, N.K. Carbon. 1999, 37, 1323-1332. 10. The properties of gases & liquids, Reid, R.C.; Prausnitz, J.M., Poling, B.E., McGraw-Hill: New York, 1998.