Elsevier

Journal of Chromatography A

Volume 1517, 29 September 2017, Pages 195-202
Journal of Chromatography A

Full length article
Van de Graaff generator for capillary electrophoresis

https://doi.org/10.1016/j.chroma.2017.08.026Get rights and content

Highlights

  • High performance capillary electrophoresis (CE) was performed using a Van de Graaff generator (VDG), a constant current source.

  • Applied voltages of >100 kV were achieved, enabling separation efficiencies of 3.5 × 106 plates that represented a 5.7 fold improvement over a conventional CE power supply.

  • A VDG can be used to deliver high voltages using a simple and low-cost setup for CE.

Abstract

A new approach for high voltage capillary electrophoresis (CE) is proposed, which replaces the standard high voltage power supply with a Van de Graaff generator, a low current power source. Because the Van de Graaff generator is a current-limited source (10 μA), potentials exceeding 100 kV can be generated for CE when the electrical resistance of the capillary is maximized. This was achieved by decreasing the capillary diameter and reducing the buffer ionic strength. Using 2 mM borate buffer and a 5 μm i.d. capillary, fluorescently labeled amino acids were separated with efficiencies up to 3.5 million plates; a 5.7 fold improvement in separation efficiency compared to a normal power supply (NPS) typically used in CE. This separation efficiency was realized using a simple set-up without significant Joule heating, making the Van de Graaff generator a promising alternative for applying the high potentials required for enhancing resolution in the separation and analysis of highly complex samples, for example mixtures of glycans.

Introduction

Capillary electrophoresis (CE) is one of the most powerful separation techniques in the field of analytical chemistry. Because CE can achieve high separation efficiency and resolution, it has been widely used for the separation of complex chemical mixtures and biological samples [1], [2], [3], [4], and has been fundamental to progress made in DNA sequencing since the early 1980s [5], [6]. The separation efficiency in CE – described by the number of theoretical plates (N) – is a function of the electrophoretic mobility of the analyte A, μA, the electric field strength, E (i.e. the applied potential, V, divided by the total capillary length, Ltot), the capillary length to the detector, LD, and the diffusion constant of the analyte, D, as demonstrated in Eq. (1):N=μAELD2DThis makes E and LD the experimental parameters most critical to enhancing separation efficiency, and the system’s ability to resolve complex mixtures. The resolution, Rs, between two analytes, A and B, scales with the square root of the applied voltage, as indicated in Eq. (2):Rs=0.177μA(μAμB2)VDwhere μB is the electrophoretic mobility of analyte B.

For the separation of extremely complex samples, the efficiency of the separation method needs to be increased, and so an increase in E is a logical approach. Practically, however, increasing E leads to an increase in current and, consequently, excessive Joule heating [7], [8]. Joule heating results from the fact that the heat generated inside the capillary by the passage of current through it can only be dissipated at the surface, causing a radial temperature gradient across the capillary that compromises separation efficiency [9]. Additionally, when the rate of heat generation exceeds the rate of dissipation, boiling of the background electrolyte (BGE) results in bubble formation, disrupting the electric field and hence aborting the separation [10], [11].

Joule heating can be minimized by decreasing the current passing through the BGE, which can be achieved by increasing the electrical resistance, for example, by reducing the ionic strength and/or the capillary diameter [12]. Optimization of the conditions is therefore also required, since reducing the ionic strength will eventually lead to lower efficiencies due to enhanced dispersion, while narrowing the capillary causes problems for detection (in terms of signal intensity) and increases the potential of clogging. Once the separation system has been optimized, more complex samples could be resolved by increasing the LD while preserving E, which thus requires the application of higher voltages.

This provides a practical limitation, as commercially available high voltage (HV) power supplies suitable for CE are typically limited to NPS, with only a few newer power supplies going up to 60 kV. To circumvent this issue, the group of Jorgenson developed an ultra-high voltage CE system based on a Cockcroft-Walton voltage multiplier design, applying voltages as high as 330 kV across a capillary just under 6 m long [13], [14], [15]. The system was applied to the separation of peptides, proteins and DNA, with peptide separations at 580 V/cm yielding separation efficiencies up to 107 plates; an order of magnitude greater than realized at the traditional applied potential of an NPS. However, this significant achievement comes at a cost in experimental complexity in order to minimize Joule heating through enhanced heat dissipation and to prevent dielectric breakdown of the capillary. During operation, the capillary was immersed in oil and a dedicated sampling interface was developed. The group of Riekkola developed a high voltage CE system capable of working at up to 60 kV [16], minimizing Joule heating by reducing the BGE conductivity using alcohols rather than aqueous buffer. Fast separations with efficiencies up to 220,000 plates were realized, but Joule heating remained problematic and the experimental setup required additional electrical insulation to prevent dielectric breakdown [17], [18], [19], [20].

Here, we propose an experimentally simpler alternative by employing a standard, inexpensive Van de Graaff (VDG) generator as a power supply [21]. A VDG is a static electricity generator that can produce extremely high voltages (350–900 kV) from the friction between a rubber belt and rollers made from two different materials. Because of their design, these high voltages can only be realized at low currents (10–60 μA), making VDG generators unattractive for most applications. For CE, however, low currents imply an inherent advantage of producing minimal Joule heating. In the work presented here, the use of a VDG as power source in CE is explored, maximizing the voltage by maintaining high electrical resistance across the capillary through optimization of the BGE ionic strength and capillary diameter. Using 2 mM borate buffer and a 90 cm long, 5 μm i.d. (inner diameter) fused silica capillary, potentials of up to 104 kV were realized, leading to separation efficiencies of up to 3,480,000 plates using a rather simple and inexpensive experimental setup.

Section snippets

Reagents

Sodium hydroxide, sodium tetraborate, fluorescein isothiocyanate isomer I (FITC) and amino acids (l-glutamic acid (Glu), l-glutamine (Gln) and l-alanine (Ala)) were purchased from Sigma-Aldrich (Germany). A 100 mM tetraborate (borax) buffer at pH 9.21 was used as a stock solution for preparing the background electrolyte (BGE). A sample stock solution containing a mixture of 5 mM of each of the 3 amino acids (AAs) was prepared in BGE. The AAs were labelled with 5 mM FITC and stored in the dark for

Results and discussion

With a conventional DC power supply, the output potential is set, and through Ohm’s law the current, I, is determined by the ratio of the applied voltage, V, and electrical resistance, R. As long as the instrument-specific current limit has not been reached, the output potential is not affected by the electrical resistance. In CE, minimizing the BGE conductivity is still relevant because reducing the current through the BGE reduces Joule heating. A VDG, in contrast, is a current source, meaning

Conclusions

An alternative high voltage capillary electrophoresis platform was developed based on the use of a Van de Graaff (VDG) generator to supply the separation potential. Previous high voltage CE techniques have required the use of complex setups or non-aqueous mobile phases in order to avoid problems with Joule heating. By moving away from a voltage supply to a current source, excessive Joule heating was avoided through the inherent current limitation. A 5 μm i.d. capillary with a 2 mM borate buffer

Acknowledgments

This report was financially supported by Korea Institute of Science and Technology Europe. Rosanne Guijt would like to acknowledge the Alexander von Humboldt Foundation for the award of a fellowship.

References (27)

  • J.W. Jorgenson et al.

    Zone electrophoresis in open-tubular glass capillaries

    Anal. Chem.

    (1981)
  • S. Terabe et al.

    Electrokinetic separations with micellar solutions and open-tubular capillaries

    Anal. Chem.

    (1984)
  • C.J. Evenhuis et al.

    Joule heating effects and the experimental determination of temperature during CE

    Electrophoresis

    (2009)
  • 1

    Centre for Regional and Rural Futures, Deakin University, Geelong, VIC 3200, Australia.

    2

    School of Earth and Environment, University of Leeds, Leeds LS2 9JT, UK.

    View full text