Portable capillary electrophoresis system with potential gradient detection for separation of DNA fragments

Yan XuWeidong QinSam F. Y. Li

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

Portable capillary electrophoresis system withpotential gradient detection for separation of DNAfragments

A portable capillary electrophoresis (CE) system with a novel potential gradient detec-tion (PGD) was utilized to separate DNA fragments. For the first time it was demon-strated that separation of DNA fragments in polymer solution could be detected by aportable CE system integrated with PGD, with a limit of detection (LOD) comparable tothat of the CE-ultraviolet (UV) method. Effects of buffer solution, sieving medium, andapplied voltage were also investigated. The portable CE-PGD system shows severalpotential advantages, such as simplicity, cost effectiveness, and miniaturization.

Keywords: Capillary electrophoresis / DNA fragment / Potential gradient detectionDOI 10.1002/elps.200410293

1 Introduction

CE, due to its automation, speed, high separation effi-ciency, and small sample and reagents requirements, hasbecome an alternative to slab-gel electrophoresis foranalysis of DNA fragments. Like slab-gel electrophoresis,CE needs to use a sieving medium to separate DNA frag-ments, because DNA fragments have similar mass-to-charge ratios and mobility in free solution [1]. Solutions ofentangled and uncross-linked polymers provide advan-tages over cross-linked gels during DNA separationby CE. Poly(ethylene oxide), polyvinylpyrrolidone (PVP),poly-N,N-dimethylacrylamide, and hydroxyethylcelluloseare important polymers that possess self-coating ability,and could prevent the tedious coating process and prob-lems associated with coating inhomogeneity, capillaryfouling, and limited shelf life [2]. CE in polymer solutionshas been successfully applied for mutation detection,genotyping, DNA sequencing, and gene expression anal-ysis [3–8].

UV-absorbance and laser-induced fluorescence (LIF)detection methods have been most commonly used inthe CE separation of DNA fragments. However, the LODof UV detection depends on the optical pathlength, whichcan translate into poor LOD for microcolumn-basedseparations. It has been reported that the concentration

LOD of DNA by UV absorbance (260 nm) is on the order of1024–1025 M [9]. Also, UV lamps tend to have limited life-time. LIF detection could provide exquisite sensitivity withvery low LOD. However, pre-sample processing (stainingor labeling) is required and if the purification product isused for further analysis, and destaining of the target isnecessary as well [10]. In addition, both UV and LIFdetection systems do not readily lend themselves todeveloping miniaturized (portable) systems, mainly be-cause of the high power consumption of UV-light sourceand consequently the difficulty of heat dissipation [11],the requirement of precise alignment of optical compo-nents, and the relative bulkiness of optical system.

Conductivity detection (CD) can be considered as anelectroanalytical technique which has the ability to detectany analyte irrespective of whether it contains an electro-active species or not. The only requirement is that themigrating analyte zones possess a conductivity that isdifferent from that of the carrier electrolyte [12]. It hassome attractive features compared to UV and LIF detec-tions [11–14]. Firstly, no labeling of the target analyte isrequired, since CD measures the bulk conductance of thesolution. Also, because the signal transduction into theelectronic domain is inherently straightforward, theequipment required for implementation of conductivitymeasurement can be much simpler and cheaper com-pared to UV and LIF detection. As a result, CD is the pre-ferable detector in portable systems. In addition, CD isamenable to small volume-based detection and couldoffer favorable sensitivity and LOD. On the other hand, thedevelopment of CD coupled with CE has been relativelyslow, mainly because the high voltage used for separationin CE would interfere with conductivity detector electron-ics [15]. However, research in this area has shown that byproper design of the detector cell, this interference can be

Correspondence: Prof. Sam F. Y. Li, Department of Chemistry,National University of Singapore, 3 Science Drive 3, 117543,Republic of SingaporeE-mail: [email protected]: 165-67791691

Abbreviations: CAPSO, 3-(cyclohexylamino)-2-hydroxy-1-pro-panesulfonic acid; CD, conductivity detection; PGD, potentialgradient detection; TAPS, N-[tris(hydroxymethyl)methyl]-3-ami-nopropanesulfonic acid; TBA1, tetrabutylammonium

Electrophoresis 2005, 26, 517–523 517

2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Nuc

leic

acid

s

518 Y. Xu et al. Electrophoresis 2005, 26, 517–523

effectively controlled [16–21]. Applications of CD are quiteextensive, including the detection of inorganic ions,organic acids, biogenic amines, saccharides, proteins,etc., separated by liquid chromatography (LC) or CE [12,22–32], as well as in the detection of oligonucleotides,such as PCR products separated by micro-reversed-phase LC and reverse-phase ion-pairing microcapillaryelectrochromatography [12–14].

Potential gradient detection (PGD) is a kind of CD. It isbased on the changes in the electric field strength be-tween migrating zones during electrophoresis, which areinversely proportional to the ionic mobilities of respectivezones. It is applicable to all charge-carrying compoundswith differences in mobility. The pioneering work of Mik-kers et al. [33, 34] had demonstrated the utility of PGD,but found poor reproducibility and time-dependentinstabilities of the conductivity signals compared to UVdetection. Since that time, PGD has been less widelyused. An improved PGD design which could be con-nected with different separation systems had been devel-oped by our group [35–37]. Application of this PGD in thedetection of inorganic ions and small organic moleculesseparated by CE has already been demonstrated [36, 37].The PGD shows simplicity and easy miniaturization com-pared to other CD modes, which often need sophisticatedelectronic design and micromanipulation with great careto position electrodes. In addition, applicability of thePGD, in cases of inorganic ions and small organic mole-cules, had shown to be satisfactory with results compa-rable to other CD detection modes [36, 37].

In this paper, a new design of the PGD detection cell wasdeveloped and connected to a portable CE separationsystem. The system was applied to the determination ofDNA fragments in polymer solutions. To the best of ourknowledge, this is the first study on the feasibility ofseparation of DNA fragments by a portable CE-PGD sys-tem with polymer solutions as sieving medium. The wholesystem, including the power supply, separation anddetection sections, was less than 5 kg in weight. Perfor-mance of CE-PGD for the separation and detection ofDNA fragments were demonstrated and compared withCE-UV. Effects of buffer solution, sieving medium andapplied field strength were also investigated.

2 Materials and methods

2.1 Chemicals

PVP(average Mr 1 300000), tetrabutylammoniumhydroxide,and lithium hydroxide were products of Aldrich (Milwaukee,WI, USA). Tris(hydroxymethyl)aminomethane (Tris) (molecu-lar biology grade) was obtained from Fisher Biotech (Fair

Lawn, NJ, USA). HaeIII digest of FX174 DNA (742 mg/mL,containing 11 DNA fragments: 72, 118, 194, 234, 271, 281,310, 603, 872, 1078, 1353 bp, respectively), L-histidine,2-(N-cyclohexylamino)ethanesulfonic acid (CHES), 3-(cyclohexylamino)-1-propanesulfonic acid (CAPS), 3-(cyclo-hexylamino)-2-hydroxy-1-propanesulfonic acid (CAPSO),N-[tris(hydroxymethyl)methyl]-3-aminopropanesulfonic acid(TAPS), and 3-(N-morpholino)propanesulfonic acid (MOPS)were bought from Sigma (St. Louis, MO, USA). The deionizedwater in the experiments was prepared by a Milli-Q system(Bedford, MA, USA). All other chemicals were of highestquality.

2.2 Portable CE-PGD and PGD cell

The portable CE-PGD system (CE Resources, Singapore,Republic of Singapore) is made up of CE-P2 electropho-resis system and the PGD detector. The CE-P2 containsan autosampler and a built-in power supplier which candeliver voltage of 230 kV to 130 kV. Both hydrodynamicand electrokinetic injection modes are available in CE-P2.The PGD detector is a separate part that is connected toCE-P2 via cable. It detects the potential changes from thePGD cell (home-made, see Fig. 1) and transmits the sig-nal to CE-P2. The whole system was controlled and datawere acquired and analyzed with Class Eleganza CEStation software (CE Resources). The system allows foron-site testing and point-of-care analysis with an internalbattery which lasts for at least 2 h after recharging.

Figure 1. Design of the PGD cell. (1) Separation capillary,(2) two platinum electrodes, (3) PVC holder fixed to capil-lary with running buffer inside, (4) small fracture at thecapillary, (5) vial with running buffer.


Electrophoresis 2005, 26, 517–523 Portable CE system 519

2.3 CE

Uncoated fused-silica capillaries of 50 mm ID and 360 mmOD (Polymicro Technologies, Phoenix, AZ, USA) wereused to carry out CE experiments. The CE-UV experi-ments were conducted on an instrument combining a CE-L1 system (CE Resources) with a Linear Instrument(Reno, NV, USA) UVIS 200 detector whose detectionwavelength was set to 260 nm. The above described CE-PGD system was used for CE-PGD experiments. The newcolumn was treated with 0.1 M NaOH, deionized water,and running buffer consecutively. It was flushed with run-ning buffer for 2 min between two consecutive runs orwhen any poor performance such as poor peak shape ornoisy baseline was observed. All solutions were filteredwith 0.20 mm Minisart (Göttingen, Germany) filters beforeuse. The HaeIII digest of FX174 DNA were 10-fold dilutedwith deionized water and hydrodynamically injected intothe capillaries by a positive pressure of 0.3 psi. Allexperiments were carried out with reverse polarity modeunder ambient temperature.

3 Results and discussion

3.1 Selection of buffer

Theoretical studies of CD [33, 38] show that, to obtaingood LODs of the charge-carrying targets, the mobility ofthe background counterion must be minimized while thedifference in effective mobilities between the analytesions and the background co-ion must be maximized. Allthe target DNA fragments were negatively charged during

electrophoresis in this experiment. Li1, histidine, Tris andtetrabutylammonium (TBA1) were tested as counterions.Results show that the mobility of Li1 (4.016 1024 cm2

V21s21 [39]) is too high to achieve acceptable LODs of theDNA fragments (data not shown). Mobilities of histidine(2.9661024 cm2V21s21), Tris (2.9561024 cm2V21s21),and TBA1 (2.0261024 cm2V21s21) [39, 40] are lowenough for the detection with PGD. However, neitherTBA1 nor histidine provided satisfactory separation andacceptable LOD (data not shown). We believe that suchresults may be caused by the formation of DNA-histidine/TBA1 complex: large ions with positively charged aminogroup, such as histidine and TBA1, may form chelateswith negatively charged DNA phosphate groups [41] dueto sterical hindrance. Therefore, Tris was chosen ascounterion in the experiment.

Similarly, we tested CHES, CAPSO, CAPS, TAPS, andMOPS as co-ions in this experiment. While CHES,CAPSO, CAPS, and TAPS show similar acceptableresults, only buffer containing MOPS as background co-ion was not suitable for the detection of the analytes.From Fig. 2 (only showing the electropherograms forCHES and MOPS as co-ions) we could see that therewere no peaks after 30 min in the buffer containing MOPS,when peaks of DNA fragments should appear. We attrib-ute the results to the DNA-buffer interaction in amine-based neutral buffer, which would change the effectivemobility of DNA fragments as reported by Stellwagen etal. [42]. MOPS (pKa 7.2) is a stronger acid compared toCHES, CAPSO, CAPS, and TAPS (pKa 9.3, 9.6, 10.4, and8.4, respectively). The formation of DNA-buffer complexand mutually interacting DNA-buffer complexes in the

Figure 2. Electrophoresis of FX174 DNA in buffers containing different co-ions. Experimental condi-tions: ID of capillary, 50 mm; length of capillary, 50 cm; separation voltage, 25 kV; detection, PGD. Buf-fer: (A) 30 mM CHES, 60 mM Tris, 2% w/v PVP, pH 9.1; (B) 30 mM MOPS, 60 mM Tris, 2% w/v PVP, pH 8.3.



comparatively neutral MOPS buffer might result in littlemobility differences between DNA fragments and the co-ion. However, the exact mechanism of how the interac-tions affect the detection of DNA fragments by PGDshould be further investigated. In this experiment, wechose CHES as co-ion in the experiment.

3.2 Influence of sieving medium

Sieving medium is one of the most important parametersin size-sieving CE since it determines both the DNA frag-ments’ migration behavior and the resolution. The sievingmedium that possesses high sieving ability, low viscosity,and dynamic coating ability is preferred. The low viscosityand dynamic coating ability are of special importance toPGD detection. To fill and replace viscous solution for aPGD cell, and the tedious processes to coat a PGD cellare almost impossible to realize, for both would destroythe detection window of a PGD cell. From previous stud-ies [43–46] on PVP as sieving medium for DNA separa-tion, we believe PVP may be a good choice: compared toother water-soluble neutral polymers with the same con-centration and molecular weight, PVP solution has amuch lower viscosity; moreover, PVP is one of the fewpolymers having the self-coating ability, because thehydrophilic carbonyl groups of PVP could form stronghydrogen bonding with the residual hydroxyl groups onthe capillary wall. It had been shown that the electro-osmotic flow (EOF) could be suppressed to a negligiblelevel when using 1% PVP solution as “dynamic coating”.

Figure 3 shows the separation of FX174 DNA in buffersolutions containing different PVP concentrations. In arunning buffer containing no PVP, DNA fragments did not

show any peaks within reasonable time (80 min), whichindicates that DNA fragments could not reach the detec-tor under the effect of the opposite-migrating EOF. Whenthe concentration of PVP increased from 1% w/v to4% w/v, resolution of the DNA fragments improved.However, higher PVP concentrations resulted in longermigration times. It takes about 63 min and 77 min,respectively, to finish the separation in buffer solutionscontaining 3% w/v and 4% w/v PVP. The results can beexplained by the fact that smaller mesh size will form inhigher concentration of polymer, which would lead toimproved resolution, and longer time for DNA to passthrough the capillary. In the meantime, in running bufferscontaining PVP higher than 2%, we could not detect thesmallest DNA fragment (72 bp), because higher PVP alsoresulted in worse baseline. We attribute the worse base-line to the increased solution viscosity, and the Jouleheating which is more prominent at relatively longseparation time. As a result, we chose 2% PVP solutionas the best polymer concentration for subsequentexperiments. Figure 4 shows the mobilities of DNA frag-ments in running buffers containing different concentra-tions of PVP. It is obvious that the mobility of each DNAfragment decreases with increased PVP concentration.The trend of each line in Fig. 4 appears to be thesame, which means the separation mechanism has notchanged.

To further test the possibility of using PVP as dynamicsieving medium in CE-PGD system, we investigated thereproducibility of the separation of FX174 DNA by CE-PGD. As shown in Table 1, the relative standard deviation(RSD) of the migration time measured for each DNA frag-ment was less than 1.5% in five runs. The RSD of peak

Figure 3. Electrophoresis of FX174 DNA in buffers containing different PVP concentration. Experi-mental conditions were same as in Fig. 2. Buffer: 30 mM CHES, 60 mM Tris, containing different w/vPVP concentrations. (A) 0%, (B) 1%, (C) 2%, (D) 3%, (E) 4%.



Figure 4. Mobility of FX174 DNA fragments in bufferscontaining different PVP concentrations. Experimentalconditions and buffer were same as in Fig. 3.

Table 1. Reproducibility of migration time and peak area(n = 5)

DNA(bp)

Meanmigrationtime (min)

RSD (%)migrationtime

RSD (%)peakarea

72 32.14 1.4 3.8118 34.78 1.5 3.5194 36.01 1.5 3.2234 36.92 0.8 3.0271/281 39.04 1.1 3.4310 40.00 1.1 2.9603 42.88 1.2 3.0872 46.30 1.4 2.81078 48.50 1.3 2.51353 49.78 1.4 2.6

area of each fragment was less than 3.8%. The resultssuggest that PVP serves well as dynamic sieving mediumin the PGD detection of DNA fragments. Also, these dataillustrated a good reproducibility of the migration time andpeak area for qualitative and quantitative analysis of DNAfragments by the CE-PGD system.

3.3 Influence of electric field strength

We performed the separation of FX174 DNA under differ-ent applied voltages (Fig. 5). As expected, increasedapplied voltage led to higher mobility and shorter analysistime. In the mean time, increased applied voltage resultedin poorer separation, which is opposite to the resultsobtained by Yan et al. [45] who achieved improvedseparation efficiency with higher voltage. However, thisresult is in agreement with the reptation theory of DNAseparation by CE in entangled polymer solution [47]. Be-cause of the induced orientation, the coiled DNA frag-ments become more elongated as the field strength isincreased. The more the migrating DNA fragmentbecomes a rod, the less the electrophoretic mobilitybecomes dependent on molecular size. Moreover, higherseparation voltage would lead to poorer baseline, be-cause PGD works under direct current (DC) mode and thevoltage supply would introduce some noise when work-ing under high voltage. However, with applied voltage lessthan 25 kV, the DNA fragments migrate too slowly to haveacceptable analysis time (more than 60 min). Therefore, inthe experiment we chose 25 kV as applied voltage for theseparation.

Figure 5. Electrophoresis of FX174 DNA under different applied voltages. Experimental conditionswere same as in Fig. 2. Separation voltage, (A) 23 kV, (B) 24 kV, (C) 25 kV, (D) 26 kV, (E) 27 kV,(F) 28 kV, (G) 29 kV, (H) 210 kV, (I) 212 kV. Buffer, 30 mM CHES, 60 mM Tris, 2% w/v PVP, pH 9.1.



Figure 6. Demonstration of separation of FX174 DNA by CE-PGD and CE-UV. Experimental condi-tions: ID of capillary, 50 mm, length of capillary, 50 cm; separation voltage, 25 kV; detection: (A) UV,(B) PGD. Buffer: (A) 30 mM CHES, 60 mM Tris, 2% w/v PVP, pH 9.1. Concentration of the DNA sample,74.2 mg/mL.

3.4 Separation performance

After evaluating different ratios and different concentra-tions of Tris and CHES, we found the buffer solution con-taining 60 mM Tris, 30 mM CHES, and 2% w/v PVP couldprovide satisfactory results for the separation ofFX174 DNA. Figure 6B shows that, except for fragments5, 6, 7 (271, 281, 310 bp), baseline separation could beachieved for other DNA fragments. Compared to thatdone with UV detector under the same condition (Fig. 6),the signal intensity of PGD is larger than UV, but withworse baseline. Consequently, the LOD of CE-PGD is al-most the same as CE-UV for the separation of DNA frag-ments. The LOD (S/N = 3) of CE-PGD for the separationof FX174 DNA, based on the smallest (72 bp) fragmentwas 79 ng/mL, while the LOD (S/N = 3) of CE-UV for thesmallest (72 bp) fragment was 91 ng/mL.

4 Concluding remarks

A novel designed PGD cell for portable CE system wasdeveloped, which shows several potential advantages,such as simplicity, cost effectiveness, and miniaturization.For the first time, it was possible to perform DNA separa-tion in polymer solution by the CE-PGD system. The LODachieved by PGD appeared to be comparable to thatachieved by UV detector.

The authors thank the National University of Singapore forfinancial support.

Received November 3, 2004

5 References

[1] Ren, J. C., Fang, N. F., Wu, D., Anal. Chim. Acta 2002, 470,129–135.

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

[3] Carrilho, E., Ruiz-Martinez, M. C., Berka, J., Smirnov, I.,Goetzinger, W., Miller, A. W., Brady, D., Karger, B. L., Anal.Chem. 1996, 68, 3305–3313.

[4] Ren, J., Ulvik, A., Refsum, H., Ueland, P. M., Clin. Chem.1998, 44, 2108–2114.

[5] Ulvik, A., Ren, J., Refsum, H., Ueland, P. M., Clin. Chem.1998, 44, 264–269.

[6] Kim, Y., Yeung, E. S., J. Chromatogr. A 1997, 781, 315–325.

[7] Gao, Q., Pang, H. M., Yeung, E. S., Electrophoresis 1999, 20,1518–1526.

[8] Mitchell, J. W., Walsh, D., Electrophoresis 1998, 19, 80–85.

[9] Singhal, P., Kuhr, W. G., Anal. Chem. 1997, 69, 3552–3557.

[10] Owens, C. V., Davidson, Y. Y., Kar, S., Soper, S. A., Anal.Chem. 1997, 69, 1256–1261.

[11] Kappes, T., Galliker, B., Schwarz, M. A., Hauser, P. C.,Trends Anal. Chem. 2001, 20, 133–139.

[12] Galloway, M., Stryjewski, W., Henry, A., Ford, S. M., Llopis,S., McCarley, R. L., Soper, S. A., Anal. Chem. 2002, 74,2407–2415.

[13] Galloway, M., Soper, S. A., Electrophoresis 2002, 23, 3760–3768.

[14] McWhorter, S., Soper, S. A., J. Chromatogr. A 2000, 883, 1–9.

[15] Zemann, A. J., Trends Anal. Chem. 2001, 20, 346–354.

[16] Zemann, A. J., Schnell, E., Volgger, D., Bonn, G. K., Anal.Chem. 1998, 70, 563–567.

[17] Schnierle, P., Kappes, T., Hauser, P. C., Anal. Chem. 1998,70, 3585–3589.

[18] Kappes, T., Schnierle, P., Hauser, P. C., Anal. Chim. Acta1999, 393, 77–82.



[19] Tanyanyiwa, J., Galliker, B., Schwarz, M. A., Hauser, P. C.,Analyst 2002, 127, 214–218.

[20] Macka, M., Hutchinson, J., Zemann, A., Zhang, S. S., Had-dad, P. R., Electrophoresis 2003, 24, 2144–2149.

[21] Unterholzner, V., Macka, M., Haddad, P. R., Zemann, A.,Analyst 2002, 127, 715–718.

[22] Meissner, T., Niess, M., J. Chromatogr. A 2004, 1035, 271–275.

[23] Yamamoto, A., Yasuhara, A., Kodama, S., Matsunaga, A.,Suzuki, S., Mohri, S., Yamada, M., J. Sep. Sci. 2004, 27,325–329.

[24] Tanyanyiwa, J., AbadVillar, E. M., Hauser, P. C., Electropho-resis 2004, 25, 903–908.

[25] Cinquina, A. L., Cali, A., Longo, F., DeSantis, L., Severoni, A.,Abballe, F., J. Chromatogr. A 2004, 1032, 73–77.

[26] Kubán, P., Kubán, V., Anal. Bioanal. Chem. 2004, 378, 378–382.

[27] Surowiec, I., Kaml, I., Kenndler, E., J. Chromatogr. A 2004,1024, 245–254.

[28] Lu, S. B., Sun, X. M., Shi, C. O., Zhang, Y. X., J. Chromatogr.A 2003, 1012, 161–168.

[29] Kubán, P., Kubán, V., Electrophoresis 2003, 24, 1935–1943.[30] Carvalho, A. Z., DaSilva, J. A. F., DoLago, C. L., Electropho-

resis 2003, 24, 2138–2143.[31] Haber, C., VanSaun, R. J., Jones, W. R., Anal. Chem. 1998,

70, 2261–2267.[32] Klampfl, C. W., Katzmayr, M. U., Buchberger, W., Electro-

phoresis 1998, 19, 2459–2464.

[33] Mikkers, F., Everaerts, F., Verheggen, T., J. Chromatogr.1979, 169, 1–10.

[34] Mikkers, F., Everaerts, F., Verheggen, T., J. Chromatogr1979, 169, 11–20.

[35] Li, S. F. Y., Wei, H. P., Zhang, G. X., PCT Int. Appl. 2000,36 pp.

[36] Feng, H.T., Wei, H.P., Li, S. F. Y., Electrophoresis 2004, 25,909–913.

[37] Qin, W. D., Li, S. F. Y., Electrophoresis 2003, 24, 2174–2179.

[38] Katzmayr, M. U., Klampl, C. W., Buchberger, W., J. Chro-matogr. A 1999, 850, 355–362.

[39] Beckers, J. F., Electrophoresis 2003, 24, 548–556.

[40] Dean, J. A., Lange’s Handbook of Chemistry, 15th Edition,McGraw-Hill, New York 1999, pp. 8157–8160.

[41] Stellwagen, N. C., Gelfi, C., Righetti, P. G., J. Chromatogr. A1999, 838, 179–189.

[42] Stellwagen, N. C., Bossi, A., Gelfi, C., Righetti, P. G., Anal.Biochem. 2000, 287, 167–175.

[43] Wang, Y. M., Liang, D. H., Hao, J. C., Fang, D. F., Chu, B.,Electrophoresis 2002, 23, 1460–1466.

[44] Gao, Q. F., Yeung, E. S., Anal. Chem. 1998, 70, 1382–1388.

[45] Yan, X. M., Hang, W., Majidi, V., Marrone, B. L., Yoshida, T.M., J. Chromatogr. A 2002, 943, 275–285.

[46] Song, L. G., Liu, T. B., Liang, D. H., Fang, D. F., Chu, B.,Electrophoresis 2001, 22, 3688–3698.

[47] Grossman, P. D., Soane, D. S., Biopolymers 1991, 31, 1221–1228.


Documents

Portable capillary electrophoresis system with potential gradient detection for separation of DNA fragments