A

irgosao©

K

1

maecoa

cCwm3ttCs

0d

Available online at www.sciencedirect.com

Journal of Chromatography A, 1185 (2008) 161–166

Microtrapping characteristics of single and multi-walled carbon nanotubes

Chaudhery Mustansar Hussain a, Chutarat Saridara b, Somenath Mitra c,∗a National Center of Excellence for Environmental and Hazardous Waste Management, Chulalongkorn University, Bangkok, Thailand

b Rajamangala University of Technology, Thanyaburi, Thailandc Department of Chemistry and Environmental Science, New Jersey Institute of Technology, Newark, NJ 07102, USA

Received 25 November 2007; received in revised form 21 January 2008; accepted 22 January 2008Available online 2 February 2008

bstract

Carbon nanotubes (CNTs) possess some highly desirable sorbent characteristics, which make them attractive for a variety of applicationsncluding micro-scale preconcentration. The main advantage of CNTs is that they are non-porous, thus eliminating the mass transfer resistanceelated to diffusion into pore structures. Their high aspects ratio leads to large specific capacity, consequently they have the potential to be the nexteneration high performance sorbent. In this paper we present the microtrapping. The objective of this paper was to study the sorption of selectrganic compounds on single and multi-walled nanotubes either packed or self-assembled onto a micro-sorbent trap. The data show that the CNTs

how highly favorable adsorption as well as desorption. The former is characterized by relatively large breakthrough volumes and isosteric heats ofdsorption (�Hs, close to 64 kJ/mol). Similarly, rapid desorption from CNTs was demonstrated by narrow desorption bandwidth. The eliminationf non-tubular carbons (NTC) from the CNT surface is important, as they reduce the performance of these sorbents. 2008 Elsevier B.V. All rights reserved.

ttrewvotta

snccc

eywords: Carbon nanotubes; Sorbent; Microtrap; Gas chromatography

. Introduction

Carbon nanotubes (CNTs) possess some highly desirableechanical, thermal and electrical properties, which make them

ttractive for a wide range of applications ranging from fieldmission to reinforcements in nanocomposites. Single-walledarbon nanotube (SWNT) and the multi-walled carbon nan-tubes (MWNT) are formed by seamless rolled up of a singlend multi-layers of graphene sheets respectively [1–3].

CNTs also have some excellent sorbent characteristics, whichan be utilized in diverse applications. Studies have shown thatNTs can efficiently remove trace contaminants from air andater [4,5], and also trap small molecules such as H2, O2 andethane. Their reported surface areas range between 150 and

000 m2/g [6], which make them attractive as sorbents. Thewo important issues in sorbent applications are high adsorp-

ion capacity and efficient desorption. The main advantage ofNTs is that they are non-porous, and the solute is held on the

urface by van der Walls type forces. This eliminates the mass

∗ Corresponding author. Tel.: +1 973 596 5611; fax: +1 973 596 3586.E-mail address: mitra@njit.edu (S. Mitra).

tmoe

sb

021-9673/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.chroma.2008.01.073

ransfer resistance related to the diffusion into the pore struc-ures. The large specific capacity comes from the high aspectatio of the CNTs. As a result, CNTs have been shown to bexcellent high resolution gas chromatography stationary phases,hich can separate a wide range of compounds from the veryolatile to relatively large polycyclic aromatics [7–11]. Basedn what has been reported [12–14], both retention and desorp-ion are highly favorable on CNTs. They have the potential to behe next generation, high performance sorbents for both volatilend semi volatile organic compounds.

The synthesis of CNTs generally involves the catalytic dis-ociation of organic precursors or graphite, during which otheron-tubular carbons (NTC), such as, amorphous and graphiticarbon are also formed [15]. Often, the NTC is formed when theatalyst is exhausted, so most preparations include the CNTsoated with NTC on the surface. Unlike the CNT, the NTCends to have porous structures, leading to a different trapping

echanisms that can be diffusion controlled [16–18]. So, purityf the CNTs is an important factor that influences the overall

ffectiveness as a sorbent.

The sorption sites on CNTs are on the wall, and in the inter-titial spaces between tubes. These sites are easily accessed foroth adsorption and rapid desorption. The NTC coverage on

1 mato

tttaap

fsucasrs

cBcFibbtbTnptSfpT[ta

sacsa

2

2

pUiasstaM

wwp

2

IvwmF5roacrp3

4MoO

2

mactptGoTa(tiCfliw

3

3

62 C.M. Hussain et al. / J. Chro

he CNT reduces their availability, as the sorbate has to diffusehrough the NTC to reach the CNT. Moreover, the porous struc-ure of NTC introduces mass transfer limitations, slowing bothdsorption and desorption [19–21]. Understanding these char-cteristics is important for their application as separation andreconcentration media.

The excellent features of CNTs, along with their nano-scaleeatures, make them ideal candidates for micro-scale devices,uch as, a microconcentrator or a sorbent trap, which have beensed in a variety of chromatography [22–24] and sensing appli-ations [25–27]. The purpose of such devices is usually to acts fast preconcentrators, or to modulate the concentration of atream for real-time monitoring [8,28–30]. They have been fab-icated in small capillary tubings and also by micromachiningilicon and other substrates [31,32].

The breakthrough and desorption efficiency are importantharacteristics of a micro-sorbent trap (referred to as microtrap).ecause of its small dimensions only a small amount of sorbentan be packed inside, thus, making it prone to breakthrough.or quantitative sampling, the sample volume should not exceed

ts breakthrough volume (BTV) defined as the volume that cane sampled per unit weight of the sorbent before the analytereaks through the sorbent bed. Previous studies have suggestedhat for a trap with a large number of theoretical plates, thereakthrough is independent of the analyte concentration [33].he microtraps are designed to be small, so that they have largeumber of plates and can be heated rapidly. Therefore, they areacked with a small amount of sorbent, often designed to retainhe analytes only for a few s/min before rapid desorption [34].o, both adsorption and desorption play important roles in aast acting microtrap, and an understanding of these processesrovides insight into the characteristics of these nano-sorbents.he CNT sorbents can be used in a microtrap in a packed format

35–37], or as a self-assembled trap [8,9]. The former facilitateshe use of a larger amount of sorbent, while the latter providesn ordered nanostructure.

The objective of this paper is to study the sorption ofelect volatile organic compounds on SWNT, MWNT and self-ssembled CNTs. The goal is to study these sorbents on aomparative basis, and evaluate their effectiveness as analyticalorbents. The other objective is to evaluate their applicability inmicro-sorbent trap.

. Experimental

.1. Materials

A MWNT sample synthesized by CVD using ethanol as arecursor and Ni as catalyst was obtained from Chaing Mainiversity [38]. This sample, which had not been standard-

zed contained significant NTC is referred to as MWNT-1. Inn effort to eliminate the NTC and purify the CNTs, first theample was passed through 106 �m (Endecott, Ltd., England)

ieve to eliminate the large particles. The NTC that remained onhe CNT surface was removed by selective oxidation in a flow ofir at 350 ◦C for 30 min. The cleaned MWNT-1 was referred asWNT-1C. Purified SWNT and MWNT (referred MWNT-2)

test

gr. A 1185 (2008) 161–166

ere purchased from Cheap Tubes Inc., CarbopackTM, whichas purchased from Supelco Inc. was used as a basis for com-arison.

.2. Microtrap fabrication

The CNTs and CarbopackTM were packed inside a 0.5 mmD capillary to form a microtrap. Mechanical shaking using aibrator was used to obtain a uniform packing. Each microtrapas packed with 20 mg of the sorbent. Self-assembled CNTicrotrap was fabricated using a procedure reported before [39].irst the steel tubing was heated in air a 10 mL/min flow at00 ◦C for 30 min to oxidize the surface. Then the surface waseduced in a 10 mL/min flow of H2 at 500 ◦C for 30 min. Thexidation and reduction led to the formation of a catalyticallyctive surface. After this, chemical vapor deposition (CVD) wasarried out at 700 ◦C for 1 h. using C2H4 as the precursor. Toemove the NTC from CNT surface, a thermal annealing in theresence of oxygen was carried out at 350 ◦C at a flow rate of00 mL/min.

Scanning electron microscopy (SEM) using Model Genesis000 XMS (EFI, USA) was used to study the MWNT-1 andWNT-1C. The SEM image of the self-assembled CNTs was

btained by a Leo 1530 VP (Carl Zeiss SMT AG Company,berkochen, Germany).

.3. Experimental system

The experimental system is shown in the Fig. 1. Parts perillion level organic compound standards were generated usingdiffusion tube method reported before [40]. The N2 stream

ontaining the organic compounds flowed through the micro-rap continuously and were retained by the sorbents. Desorptionulses at fixed intervals were applied to the microtrap, so that therapped organic compounds were desorbed and detected by theC. The microtrap was resistively heated with a 7–10 A pulsef electric current from a power supply (Variac 100/200 Series).ypical duration of the electric pulse was between 1 and 1.5 s,nd intervals varied between 2 and 20 min. An electric timerVariac ATC 305) was used to control the frequency of injec-ions. Different desorption temperatures could be achieved byncreasing either the voltage, or the duration of heating. Gashromatograph (Varian 3400) equipped with a conventionalame ionization (FID) was used for analysis using a 0.53 mm

.d., 30 m capillary column (DB-624, J&W Scientific). Nitrogenas used as the carrier gas at a flow rate of 3 mL/min.

. Results and discussion

.1. Purifications of CNTs

Fig. 2(a) shows the SEM image of MWNT-1, which appears

o be covered with NTC. Some of it comprised of large agglom-rates of several microns in diameter. With few available CNTites, much of the adsorption would be expected to occur onhe NTC. Different purification strategies were employed to

C.M. Hussain et al. / J. Chromatogr. A 1185 (2008) 161–166 163

m of t

chtcwthtd

FrC

w2

icassC

Fig. 1. Schematic diagra

lean MWNT-1. Fig. 2(b) is the SEM of MWNT-1C, whichad been cleaned by sieving to remove the large particles, andhen by oxidation in air at 350 ◦C to eliminate the amorphousarbon on the surface. This preparation showed mainly CNTsith some residual NTC. Too long an exposure to air at high

emperature led to the oxidation of the MWNT as well, so theeating time was limited to 30 min. With these two purifica-ion steps, the availability of the active CNT sites increasedramatically.

The SEM images of the MWNT-2 and SWNT are presented inig. 2(c) and (d) respectively, which show relatively pure prepa-ations that did not require any preprocessing. The length of theNTs ranged from a few hundred nanometers to micrometers

d

sT

Fig. 2. SEM images of (a) MWNT-1, (b) MWNT-1C, (c) M

he experimental system.

hile the average diameter ranged of 10–20 nm for SWNT and0–40 nm for MWNTs respectively.

Fig. 2(e) shows the SEM image of self-assembled CNTsnside the microtrap. This image was taken after the oxidativeleaning by passing air at 350 ◦C. The CNTs were verticallyligned in a forest-like structure providing excellent sorptionites. The average diameter for self-assembled CNTs was muchmaller than MWNT-1C. The film thickness of self-assembledNTs was found to be about 20 �m, and they were uniformly

istributed through out the microtrap.

The characteristics of the different CNTs in terms of theirurface area, average diameter and length are presented inable 1. SWNT while MWNT-2 had the highest BET (Brunauer,

WNT-2, (d) SWNT and (e) self-assembled CNTs.

164 C.M. Hussain et al. / J. Chromatogr. A 1185 (2008) 161–166

Table 1Characteristics of the different carbon nanotubes used in the microtraps

Type of sorbent Surface area(m2/g)

Averagediameter (nm)

Averagelength (�m)

MWNT-1C 110 150–175 More than 10MWNT-2 233 20–40 10–30SWNT 407 10–20 5–30SC

Et

3

omittwTfltflftltbr

ipswp

Fm

Fm

tbcgsotiosir

ftTdtttl

elf-assembled CNTs – 20–50 10–20arbopack 100 – –

mmett, and Teller) surface area that were significantly higherhan CarbopackTM.

.2. Breakthrough characteristics of the microtrap

As the N2 continuously flowed through the microtrap, therganic compounds already trapped in the microtrap began toigrate because the N2 acted as an eluent. The sorption capac-

ty of the CNTs in the microtrap was evaluated by studyinghe breakthrough time of the microtrap, which is defined as theime required by an analyte to elute through, or the time forhich the organic compounds are retained on the CNTs [41].he breakthrough volume is a product of breakthrough time andow rate, so it is directly proportional to the latter. The break-

hrough time is a function of the capacity factor, length, andow rate [8]. It was measured by operating the microtrap at dif-erent injection intervals. Increasing injection interval enhancedhe detector response as more organic compounds were accumu-ated by the microtrap. Once the VOCs began to breakthrough,he response did not increase any further with the interval. So thereakthrough time of the microtrap was computed as the timeequires to reach the maximum response [42,43].

Five organic compounds represented a wide range of volatil-ty with varying polarity were analyzed. Typical breakthrough

◦

rofiles for SWNT at 25 C are seen in Fig. 3, with toluenehowing the longest breakthrough time. Sorption on other CNTsere similar, and are not presented here for brevity. The com-arison of different sorbents when it comes to the retention of

ig. 3. Detector response as a function of injection interval for the SWNTicrotrap.

aoxtbxTdsbt

TB

T

MMSMCS

ig. 4. Detector response as a function of injection interval for toluene usingicrotraps containing different sorbents.

oluene at 25 ◦C is shown in Fig. 4. For the same mass of sor-ent, all the CNTs showed significant higher sorption capacityompared to Carbopack, the only conventional sorbent in theroup. The impure MWNT-1 with large amount of NTC wasimilar to the Carbopack as much of the adsorption occurredn the NTC, which was morphologically closer to Carbopackhan the CNTs. The purification to MWNT-1C led to a 600%ncrease in breakthrough time. This demonstrates the importancef impurity elimination when it comes to the use of CNTs. Theelf-assembled CNTs showed relatively high sorption capac-ty in spite of the fact that it was only a 20 �m thick film andepresented small quantity of CNTs.

The breakthrough times for different compounds on the dif-erent sorbents at 25 ◦C are listed in Table 2. The SWNT showedhe highest adsorption capacity in terms of breakthrough volume.his is attributed to the higher aspect ratio due to the smalleriameter, which led to the higher specific area. For example,he breakthrough time of DCM on SWNT was five times higherhan that on Carbopack or MWNT-1. MWNT-1C showed break-hrough time that were closer to MWNT-2, but were consistentlyower. This was attributed to higher impurities in the former,long with other factors, such as, different size and morphol-gy. The retention of C6H6 was lower than expected, but that ofylene was quite high. The aromatic compounds were expectedo have relatively stronger interactions with CNTs [8]. Althoughenzene and hexane exhibited similar breakthrough times, whileylene showed significantly longer breakthrough times. Fromables 1 and 2 it appears that the breakthrough times were notirectly related to the BET surface area. For example, the specific

urface area of SWNT was significantly higher than MWNT-2,ut for many of the compounds studied here, the breakthroughimes were quite similar.

able 2reakthrough times (min) for different sorbents and analytes at 25 ◦C

ype of microtrap DCM Ethanol Hexane Benzene Toluene Xylene

WNT-1C 5 10 15 15 40 120WNT-2 8 12 20 20 45 120

WNT 10 15 20 20 50 120WNT-1 2 2 5 6 6 10arbopack 2 2 5 6 6 10elf-assembled CNTs 2 5 6 10 15 20

C.M. Hussain et al. / J. Chromatogr. A 1185 (2008) 161–166 165

F

3

ttdfFt

l

wsts

rtTag

�

vctoiaTmiM�

ttta

Table 3Isosteric heats of adsorption (�Hs) for different carbon nanotubes using benzeneas the analyte

Sorbents �Hs (kJ/mol)

MWNT-1 18.49MWNT-1C 44.92MWNT-2 56.43SWNT 63.91Carbopack 18.155

FS

3

ttedpimportant parameter in chromatographic separation.

Table 4 provides an interesting insight into the desorptionof toluene and benzene from all the sorbents. Due to the masstransfer issues related to porous structures, both MWNT-1 and

Table 4Peak width at half height for toluene and benzene

Type of microtrap Toluenea (s) Benzeneb (s)

Self-assembled CNTs 1.88 0.275MWNT-1C 1.87 0.28MWNT-2 1.7 0.265SWNT 1.6 0.26

ig. 5. Breakthrough time as a function of 1/T. Benzene was used as the analyte.

.3. Breakthrough time as a function of temperature

Adsorption is an exothermic phenomenon and the break-hrough time of a compound is inversely proportional to theemperature. As expected when adsorption temperature wasecreased the breakthrough time increased [44]. The resultsollowed the Van’t Hoff-type relationship [45–47] as shown inig. 5. The plot of log BTV as a function of 1/T and was found

o be linear according to:

og(BTV) = k11

T+ k2 (1)

here, BTV is the breakthrough volume and k1 and k2 are con-tants. T is the absolute temperature. It was interesting to notehat the slopes varied for the different sorbents with SWNThowing the highest while Carbopack the lowest.

The isosteric heat of adsorption, �Hs is the amount of heateleased when an atom adsorbs on a substrate, and is related tohe activation energy of sorption for a sorbate–sorbent system.he strength of interaction of compound with the surface of thedsorbent is represented by the enthalpy of adsorption, �Hs,iven by [45,48]

Hs = −R∂(ln BTV)

∂(1/T )(2)

The �Hs were obtained from the slope of plots of ln BTVs. 1/T, where BTV is the breakthrough volume of the organicompound on the sorbent, T is the absolute temperature and R ishe gas constant. A linear dependence indicated a constant valuef the isosteric heat of adsorption in the temperature range stud-ed, while relative change in �Hs of sorbents with temperature isttributable to the activation of the sorbent surface [9,44,48–52].hese values for benzene are presented in Table 3. The maxi-um �Hs was for SWNT, suggesting that it had the strongest

nteraction with the analyte. This was followed by MWNT-2 andWNT-1C. MWNT-1 and Carbopack showed similar values ofHs, about a third that of SWNT. Once again this demonstrated

hat the mechanisms of adsorption were quite similar in these

wo sorbents. It is interesting to note that the �Hs followed arend similar to that of the breakthrough times for the differentnalytes.

MC

ig. 6. Desorption profile of toluene with different sorbents inside microtrap (a)WNT (b) Carbopack.

.4. Quantitative desorption from the microtrap

The desorption of organic compound is also an important fac-or. It was achieved by passing a pulse of electric current directlyhrough the wall of the microtrap. For quantitative desorption,nough energy needs to be provided [53]. Fig. 6 shows typicalesorption profile of toluene generated from these sorbents. Theeak width is a measure of the ease of desorption, and is also an

WNT-1 2.05 1.87arbopack 2.06 1.89

a Flow rate of 3 mL/min.b Flow rate of 6 mL/min.

1 mato

CrfHtfatc

4

amacSnorwa

R

[

[

[[

[

[[

[[

[[

[[

[[[[

[[

[

[

[[[[[[

[[[[[[[[

[

[[

[[

66 C.M. Hussain et al. / J. Chro

arbopack showed the widest desorption bands, implying slowelease of solutes. The desorption peak widths were lowestor SWNT. For toluene at 3 mL/min, the difference was 29%.owever for benzene at a 6 mL/min, this difference increased

o more than 3.5 times. The very low desorption bandwidthsrom SWNT, MWNT-2 and MWNT-1C prove that the CNTsre relatively easy to desorb and should provide higher desorp-ion efficiency. Overall, the CNTs not only have high adsorptionapacity but also are desorbed easily.

. Conclusions

The carbon nanotubes are excellent for both strong adsorptionnd relatively rapid desorption of organic compounds. The for-er is characterized by high breakthrough times (or volumes)

nd isosteric heats of adsorption (�Hs) close to 64 kJ/mol asompared to 18 kJ/mol for a non-tubular carbon studied here.imilarly, rapid desorption from CNTs was demonstrated byarrow desorption bandwidths. The elimination of NTC and dis-rdered carbon from the CNT surface is quite important, as theyeduce the performance of these sorbents. Both single and multi-alled tubes showed excellent performance during adsorption

s well as desorption.

eferences

[1] S. Iijima, Nature 56 (1991) 354.[2] T.W. Ebbesen, H.J. Lezec, H. Hiura, J.W. Bennett, H.F. Ghaemi, T. Thio,

Nature 54 (1996) 382.[3] T.P. Kumar, R. Ramesh, Y.Y. Lin, G.T. Fey, Electrochem. Commun. 6

(2004) 520.[4] Y.H. Li, S. Wang, J. Wei, X. Zhang, C. Xu, Z. Luan, D. Wu, B. Wei, Chem.

Phys. Lett. 205 (2001) 336.[5] F.L. Dark, P. Malbrunot, G.P. Tartaglia, Int. J. Hydrogen Energy 2 (2002)

193.[6] X.Y. Zhu, S.M. Lee, Y.H. Lee, T. Frauenheim, Phys. Rev. Lett. 85 (2000)

2757.[7] A. Lan, A. Mukasyan, J. Phys. Chem. B 109 (2005) 16011.[8] C. Saridara, R. Brukh, Z. Iqbal, S. Mitra, Anal. Chem. 77 (2005) 1183.[9] M. Karwa, S. Mitra, Anal. Chem. 78 (2006) 2064.10] M. Stadermann, A.D. McBrady, B. Dick, V.R. Reid, A. Noy, R.E. Synovec,

O. Bakajin, Anal. Chem. 78 (2006) 5639.11] L.M. Yuan, C.X. Ren, L. Li, P. Ai, Z.H. Yan, M. Zi, Z.Y. Li, Anal. Chem.

78 (2006) 6384.12] H. Hasar, J. Hazard. Mater. B 97 (2003) 49.13] M. Bienfait, P. Zeppenfeld, N. Dupont-Pavlovsky, M. Muris, M.R. Johnson,

T. Wilson, M. DePies, O.E. Vilches, Phys. Rev. B 70 (2004) 5410.14] H. Li, S. Wang, Z. Luan, J. Ding, C. Xu, D. Wu, Carbon 41 (2003) 1057.

[[[[

gr. A 1185 (2008) 161–166

15] R. Brukh, S. Mitra, Chem. Phys. Lett. 424 (2006) 126.16] S. Maruyama, R. Kojima, Y. Miyauchi, S. Chiashi, M. Kohno, Chem. Phys.

Lett. 360 (2002) 229.17] X. Zhang, W. Wang, Fluid Phase Equilib. 194 (2002) 289.18] H.J. Rafson, Odor and VOCs Control Handbook, McGraw-Hill, New York,

1998.19] P. Liang, Y. Liu, L. Guo, Spectrochim. Acta, Part B 60 (2005) 125.20] X. Peng, J. Ding, Z. Di, Y. Li, Z. Luan, B. Tian, Mater. Lett. 59 (2005)

399.21] C. Lu, Y.L. Chung, K.F. Chang, Water Res. 39 (2005) 1183.22] G. Lambertus, A. Elstro, K. Sensenig, J. Potkay, M. Agah, S.D. Scheuering,

K. Wise, F. Dorman, R. Sacks, Anal. Chem. 76 (2004) 2629.23] G. Lambertus, R. Sacks, Anal. Chem. 77 (2005) 2078.24] J.M. Sanchez, R.D. Sacks, Anal. Chem. 78 (2006) 3046.25] A.E. Herr, A.K. Singh, Anal. Chem. 77 (2005) 585.26] A.E. Herr, D.J. Throckmorton, A.A. Davenport, A.K. Singh, Anal. Chem.

77 (2005) 585.27] S. Song, A.K. Singh, Anal. Bioanal. Chem. 384 (2006) 41.28] R. Kottenstette, R.P. Manginell, R.C. Hughes, M.W. Jenkins, S.V. Patel,

U.S. Patent No. 6902701 (2005).29] J.H. Flemming, H.K. Baca, M. Werner-Washburne, S.M. Brozik, G.P.

Lopez, Sens. Actuators B 113 (2006) 376.30] P.R. Lewis, R.P. Manginell, D.R. Adkins, R.J. Kottenstette, D.R. Wheeler,

S.S. Sokolowski, D.E. Trudell, J.E. Byrnes, M.J. Okandan, M. Bauer, R.G.Manley, G.C. Frye-Mason, IEEE Sens. J. 6 (2006) 784.

31] S. Reidy, G. Lambertus, J. Reece, R. Sacks, Anal. Chem. 78 (2006) 2623.32] S. Reidy, G. Lambertus, J. Reece, R. Sacks, Anal. Chem. 79 (2007) 2911.33] P. Lovkvist, J.A. Jonsson, Anal. Chem. 59 (1987) 818.34] P. Lovkvist, J.A. Jonsson, J. Chromatogr. A 365 (1986) 5.35] Q. Li, D. Yuan, J. Chromatogr. 100 (2003) 203.36] M. Valcarcel, B.M. Simonet, S. Cardenas, B. Suarez, Anal. Bioanal. Chem.

382 (2005) 1783.37] Q. Zhou, Y. Ding, J. Xiao, Chromatographia 65 (2006) 25.38] P. Singjai, S. Changsarn, S. Thongtem, Mater. Sci. Eng. A 443 (2007) 42.39] C. Saridara, S. Mitra, Anal. Chem. 77 (2005) 7094.40] A.C. Savitsky, S. Sigga, Anal. Chem. 44 (1972) 1712.41] C. Feng, S. Mitra, J. Chromatogr. A 805 (1998) 169.42] M. Karwa, Z. Iqbal, S. Mitra, Carbon 44 (2005) 1235.43] S. Mitra, C. Yun, J. Chromatogr. 648 (1993) 415.44] C. Thammakhet, P. Thavarungkul, R. Brukh, S. Mitra, P. Kanatharana, J.

Chromatogr. A 1072 (2005) 243.45] V. Simon, A. Riba, M.L. Waldhart, L. Torres, J. Chromatogr. A 704 (1995)

465.46] P.W. Atkin, Physical Chemistry, Oxford University Press, Oxford, 1994.47] P. Safaeefar, H.M. Ang, H. Kuramochi, Y. Asakuma, K. Maeda, M.O. Tade,

K. Fukui, Fluid Phase Equilib. 238 (2005) 180.48] E. Dıaz, S. Ordonez, A. Vega, J. Colloid Interface Sci. 305 (2007) 7.49] D.H. Yoo, G.H. Rue, Y.H. Hwang, H.K. Kim, Phys. Chem. B 106 (2002)

3371.50] C. Bilgic, A. Askin, J. Chromatogr. A 1006 (2003) 281.51] K. Yang, L. Zhu, B. Xing, Environ. Sci. Technol. 40 (2006) 1855.52] C.H. Wu, J. Colloid Interface Sci. 311 (2007) 338.53] S. Mitra, Y.H. Xu, W. Chen, A. Lai, J. Chromatogr. A 727 (1996) 111.