Yan XuSam Fong Yau Li

Department of Chemistry,National University of Singapore,Republic of Singapore

Received April 3, 2006Revised April 30, 2006Accepted May 4, 2006

Short Communication

Carbon nanotube-enhanced separation ofDNA fragments by a portable capillaryelectrophoresis system with contactlessconductivity detection

It was demonstrated that separation of DNA fragments by a CE-contactless con-ductivity detection system (CE-CCD) could be enhanced with multiple-wall carbonnanotubes (MWCNs) as buffer additive. For HaeIII digest of FX174 DNA, optimizedMWCN concentration was obtained when the MWCN was above its threshold con-centration, at which MWCN could form a network in the buffer as pseudostationaryphase to provide additional interaction sites. In the case of larger DNA, MWCN near orbelow its threshold concentration was enough to provide great improvement of theresolution, which was shown by the separation of the 2-Log DNA ladder. Furthermore,the buffer containing MWCN could provide a more stable baseline in the CE-CCDsystem, owing to less fluctuation of its conductivity. Compared with CE-UV, CE-CCDwith MWCN could provide lower LODs as well as better resolution.

Keywords: Capillary electrophoresis / Carbon nanotubes / Contactless conductivitydetection / DNA fragment DOI 10.1002/elps.200600270

Among different strategies to improve the separationperformance of CE, nanostructures are attractingincreasing attention due to their unique properties.Nanoparticles could serve as capillary coatings [1, 2] orbuffer additives [3, 4], and result in enhanced CE perfor-mance. While many nanoparticle types could be used inCE separation, gold nanoparticles have been most widelyinvestigated due to their ready availability and relativestability [5, 6]. Carbon nanotubes (CNs) could be regard-ed as an ultimate fiber formed of perfectly graphitizedclosed seamless shells [7]. Since the discovery of CNs in1991 by Ijima [8], the tube-like material has drawn intenseresearch interest owing to its unique structural, mechan-ical, and electronic properties [9]. However, application ofCNs in CE is scarce, with only two relevant reports to date[10, 11]: Wang and co-workers chose CNs as the bufferadditive to improve the separation of homologs and iso-mers, and Luong et al. employed CN-coated capillary forthe separation of aniline derivatives.

There are difficulties in using CNs as CE buffer additive:the hydrophobic CN is difficult to disperse in aqueous CEbuffer, and to some extent, the black color of CN wouldinterfere with UV and LIF detection, which is the mostcommonly available detector in CE system. An effectivemethod to make CNs soluble in aqueous buffer is to treatCNs chemically to functionalize the surface [10, 11]. Useof conductivity detection (CD) as an alternative detectorcould avoid the baseline noise from the black color of CN,because CD is based on electronic instead of opticaltechnique. CNs would form polymer-like network in buffer[10] and could bind to DNA to form a DNA–CN hybrid [11,12], which prompted us to further consider that CN-con-taining buffer could afford additional interaction sites forDNA separation. In this paper, we shall demonstrate theseparation of DNA fragments by a portable CE systemwith contactless conductivity detection (CCD, CEResources, Singapore), during which multiple-wall car-bon nanotubes (MWCNs) would be investigated as bufferadditive.

A MWCN, which has been treated chemically with wetoxidation in concentrated acids, was negatively charged[13] and could be suspended in deionized water stably.After a buffer containing a polymer was prepared, a pre-determined amount of MWCN stock solution was addedinto the buffer, followed by magnetic stirring for at least60 min to obtain a homogeneous suspension. As shown

Correspondence: Professor Sam Fong Yau Li, Department ofChemistry, National University of Singapore, 3 Science Drive 3, Sin-gapore 117543, Republic of SingaporeE-mail: chmlifys@nus.edu.sgFax: 165-67791691

Abbreviations: CCD, contactless conductivity detection; MWCN,multiple-wall carbon nanotube; TEM, transmission electron micros-copy

Electrophoresis 2006, 27, 4025–4028 4025

© 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

4026 Y. Xu and S. F. Y. Li Electrophoresis 2006, 27, 4025–4028

by transmission electron microscopy (TEM, JEOL-2010,JEOL, Japan), MWCNs tend to form polymer-like net-works in both deionized water and polymer solution(Fig. 1).

Figure 2 shows that the mobilities of FX174 DNA frag-ments depend on the concentration of MWCNs. At lowconcentration, the mobilities changed little; with 10 ppm

MWCNs, the DNA fragments’ mobilities were almost thesame as those without MWCNs. Addition of 20 ppmMWCNs began to change the mobility slightly. When theMWCN concentration reached a certain level (25 ppm inthis case) each additional increase of the MWCN con-centration could cause a significant rise in the mobility. Inour opinion, in the case of low concentration, the nega-tively charged MWCNs could form an alignment along the

Figure 1. TEM of MWCN. (A)100 ppm MWCN in deionizedwater. (B) 25 ppm MWCN in2% w/v PVP solution.

Figure 2. Mobility and electro-phoresis of FX174 DNA by CE-CCD in buffers containing differ-ent concentrations of MWCN.Experimental conditions: capil-lary, 50 mm id, 360 od mm,60 cm long with a detector win-dow at 55 cm from the inlet;injection, 25 kV for 5 s; separa-tion voltage, 28 kV; detection,CCD. Buffer: 30 mM CHES,60 mM Tris, 2% w/v PVP con-taining different concentration ofMWCN: (A) 0 ppm, (B) 10 ppm,(C) 20 ppm, (D) 25 ppm,(E) 30 ppm, (F) 40 ppm, and(G) 50 ppm. 1–11: 72, 118, 194,234, 271, 281, 310, 603, 872,1078, and 1353 bp, respec-tively. Concentration of the11 DNA fragments: 25.6 mg/mL.

© 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

Electrophoresis 2006, 27, 4025–4028 CE and CEC 4027

applied electric field and migrate towards the anode [14],which would have little effect on the DNA fragments’ mo-bilities. When the MWCNs reach a certain concentration,a network would form as shown in Fig. 1, which is similarto the behavior of a polymer above its entanglementthreshold. Therefore, in addition to the network formed byPVP, there is another network in the buffer formed byMWCNs, which could serve as a pseudostationary phaseto afford additional interaction sites and change the DNAfragments’ mobilities.

However, the resolution will not always improve withincreased MWCNs. Electrophoresis in Fig. 2 shows that ina running buffer containing no MWCNs, DNA frag-ments 5–7 could not be baseline-separated. Resolutionimproved with increased MWCNs up to 25 ppm; furtherincreases in MWCN concentration resulted in decreasedresolution. In our opinion, the negatively charged networkformed by MWCNs could offer not only sieving abilitywhich would favor the DNA separation, but also electro-static repulsion of the negatively charged DNA whichwould change the mobility. At higher MWCN concentra-tions, the repulsive effect would dominate, and result infaster separation but poorer resolution. In addition, theformation of DNA–MWCN hybrid under different MWCNconcentrations might change the DNA fragments’ mobili-ties and affect the resolution. In our experiment, 25 ppm isthe optimized MWCN concentration for the separation ofFX174 DNA fragments.

Figure 3 shows the electrophoresis of FX174 DNA by CE-CCD with and without MWCNs. It is apparent that thebuffer with 25 ppm MWCNs could provide a more stablebaseline. We measured the conductivity of the bufferwithout and with 25 ppm MWCNs (MC126-2M portableconductivity meter, Mettler-Toledo, reference tempera-ture: 257C). Insert of Fig. 3 shows the average con-ductivity values (n = 3) over the temperature range of 20–357C. It was observed that the fluctuation of conductivity

was less in the buffer containing 25 ppm MWCNs thanthat without MWCNs, which could explain the morestable baseline in the buffer with 25 ppm MWCNs. Also,the formation of DNA-MWCN hybrid, which may havedifferent conductivity from free DNA fragment, mightresult in a more stable baseline in the CE-CCD system.

Compared to HaeIII digest of FX174 DNA fragments,2-Log DNA ladder contains DNA fragments of larger basepairs and larger base pair differences between each other.However, the separation was unsuccessful in the buffersolution containing 1% w/v PVP, as shown in Fig. 4A.Addition of MWCNs could greatly improve the resolutionand the baseline, with optimized MWCN concentration of15 ppm, as shown in Fig. 4B. The sieving ability offered bythe relatively low concentration of MWCNs was evidentfor the separation of the larger DNA fragments in 2-LogDNA ladder. In the meantime, the mobility changes of thelarger DNA fragments were slight, as could be seen fromthe relatively small change of the migration time. As aresult, in the case of larger DNA fragments, sieving abilityis the dominating effect from MWCNs, even at low MWCNconcentration near or below the threshold concentrationto form network. We considered this may be caused bythe fact that large DNA could drag along the MWCNmolecules it encountered during migration, which is simi-lar with the behavior of large DNA separation at polymersolution below its entanglement threshold [3]. Therefore,for the separation of larger DNA fragments by CE-CCD,we could use buffer containing low polymer concentrationwith enhancement by addition of low concentration ofMWCNs.

Table 1 lists the LODs (S/N = 3) of CE-CCD without andwith MWCNs. Compared with the LOD of CE-UV [15], theLOD of CE-CCD without MWCNs is poorer, but the LODof CE-CCD with MWCNs is better. The better LOD of CE-CCD with MWCNs can be attributed to the fact that thebuffer with MWCNs could provide a more stable baseline.

Figure 3. Electrophoresis ofFX174 DNA by CE-CCD withand without MWCN. Experi-mental conditions were thesame as those in Fig. 2. Buffer:(A) 30 mM CHES, 60 mM Tris,2% w/v PVP; (B) 30 mM CHES,60 mM Tris, 2% w/v PVP,25 ppm MWCNs. Insert: con-ductivity of buffers A and B over20–357C. 1–11: 72, 118, 194,234, 271, 281, 310, 603, 872,1078, and 1353 bp, respec-tively. Concentration of the11 DNA fragments: 25.6 mg/mL.

© 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

4028 Y. Xu and S. F. Y. Li Electrophoresis 2006, 27, 4025–4028

Figure 4. Electrophoresis of 2-Log DNA ladder by MWCN-enhanced CE-CCD. Experimental condi-tions were the same as those in Fig. 2. Buffer: (A) 30 mM CHES, 60 mM Tris, 1% w/v PVP; (B) 30 mMCHES, 60 mM Tris, 1% w/v PVP, 15 ppm MWCNs. 1–20: 100, 200, 300, 400, 500, 517, 600, 700, 800,900, 1000, 1200, 1517, 2017, 3001, 4001, 5001, 6001, 8001, and 10 002 bp, respectively. Con-centration of the 20 DNA fragments: 50.0 mg/mL.

Table 1. LODs of CE-CCD without and with MWCNs

DNA fragments CE-CCD withoutMWCNs (ng/mL)

CE-CCD withMWCNs (ng/mL)

FX174 DNAfragmentsa)

109 23c)

2-Log DNA ladderb) 67 28d)

a) LOD based on the smallest fragment (72 bp).b) LOD based on the smallest fragment (100 bp).c) LOD in buffer with 25 ppm MWCNs.d) LOD in buffer with 15 ppm MWCNs.

For the buffers without MWCNs, the LOD of FX174 DNAfragments is poorer than that of 2-Log DNA ladder, whichmay be caused by the higher PVP concentration used forFX174 DNA fragments. Thus, we could deduce thathigher polymer concentration would lead to higher vis-cosity and worse LOD in CE-CCD. Concerning the reso-lution, the buffer with 25 ppm MWCNs could provide abaseline separation for the fragments 5, 6, and 7 in FX174DNA, while both CE-CCD without MWCNs and CE-UVcould not [15]. For 2-Log DNA ladder, all the 20 DNAfragments were baseline-separated in buffer with 15 ppmMWCNs, without increasing the polymer concentrationwhich may cause worse LOD otherwise. In conclusion,MWCNs could be a complementary buffer additive topolymer for the separation of DNA fragments by CE-CCD.

The authors thank the National University of Singapore forfinancial support.

References

[1] Huber, C. G., Premstaller, A., Kleindienst, G., J. Chromatogr.A 1999, 849, 175–189.

[2] Pumera, M., Wang, J., Grushka, E., Polsky, R., Anal. Chem.2001, 73, 5625–5628.

[3] Huang, M. F., Huang, C. C., Chang, H. T., Electrophoresis2003, 24, 2896–2902.

[4] Viberg, P., Jornten-Karlsson, M., Petersson, P., Spegel, P.,Nilsson, S., Anal. Chem. 2002, 74, 4595–4601.

[5] Chiou, S. H., Huang, M. F., Chang, H. T., Electrophoresis2004, 25, 2186–2192.

[6] Huang, M. F., Kuo, Y. C., Huang, C. C., Chang, H. T., Anal.Chem. 2004, 76, 192–196.

[7] Ebbesen, T. W. (Ed.), Carbon Nanotubes, Preparation andProperties, CRC Press, Boca Raton, FL 1997.

[8] Ijima, S., Nature 1991, 354, 56–58.[9] Benedek, G., Milani, P., Ralchenko, V. G. (Eds.), Nano-

stuctured Carbon for Advanced Applications, Kluwer Aca-demic Publishers Amsterdam, The Netherlands 2001.

[10] Wang, Z. H., Luo, G. A., Chen, J. F., Xiao, S., Wang, Y.,Electrophoresis 2003, 24, 4181–4188.

[11] Luong, J. H. T., Bouvrette, P., Liu, Y. L., Yang, D.-Q., Sacher,E., J. Chromatogr. A 2005, 1074, 187–194.

[12] Zheng, M., Jagota, A., Semke, E. D., Diner, B. A. et al., Nat.Mater. 2003, 2, 338–342.

[13] Chen, G., Zhang, L., Wang, J., Talanta 2004, 64, 1018–1023.[14] Doorn, S. K., Fields, R. E., Hu, H., Hamon, M. A. et al., J. Am.

Chem. Soc. 2002, 124, 3169–3174.[15] Xu, Y., Qin, W. D., Li, S. F. Y., Electrophoresis 2005, 26, 517–

523.

© 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com