Nanoimprint lithography for the fabrication of DNA electrophoresis chips

https://doi.org/10.1016/S0167-9317(02)00511-7Get rights and content

Abstract

We describe the fabrication of microfluidic devices for bio-molecule separation using an array of well-defined nanostructures. Two types of pattern replication of the same device configuration are considered, based on different material processing. In the first approach we use a tri-layer nanoimprint lithography process to pattern a silicon dioxide substrate, on top of which we stick a transparent elastomer cover plate. The second approach relies on direct imprinting of thermoplastic polymer pellets to form two bulk plastic plates later assembled together by thermal bonding. As a result, novel microfluidic devices combining deep and wide channels and a shallower nanostructure array are obtained. The fabricated devices have been characterized by epifluorescence microscopy, using a fluorescein solution to track fluid penetration inside the high density nanostructured region. These realisations not only demonstrate that nanofluidic devices are achievable, but also that they can be manufactured for mass production via nanoimprint-based techniques.

Introduction

The Human Genome Project and human polymorphism analysis have led to much interest being devoted to the development of microfabrication technologies for more efficient DNA molecule separation [1]. One of the challenges is to introduce high density nanopillar arrays inside microchannels, to serve as artificial gels in integrated capillary electrophoresis chips. The required array area is typically several millimeters long and covers the total width of the channel, generally around 100 μm wide. Although electron beam lithography can be used to produce such nanostructure areas, a lower-cost replication technology is needed for mass production.

In this work, we present two approaches based on nanoimprint lithography [2] for the fabrication, in both silicon dioxide and plastic substrates, of capillary electrophoresis chips containing an in-channel array of 150-nm diameter pillars of period 320 nm for DNA separation. The size, shape, and period of these nanopillars can be easily changed. In the present work we chose a typical ‘pore’ size of 170 nm for demonstration. In the first approach, tri-layer nanoimprint lithography [3] was used to create the structure in a SiO2/Si substrate while soft lithography [4] was used to cast a polydimethylsiloxane (PDMS) cover sheet. In the second approach, the whole structure was created by direct imprinting of polymethylmethacrylate (PMMA) sheets, allowing a much faster and lower-cost processing of disposable microfluidic devices.

Section snippets

Device concept

A top view of the designed structure is schematically represented in Fig. 1. It contains a standard cross-shaped channel of width 50 μm ending with 500-μm diameter reservoirs. The nanostructured portion of the separation channel is 7 mm long and contains a 320-nm period triangular array of 150-nm diameter pillars. The device is made of a top and bottom part later aligned and assembled together. One part contains exactly the structure described in Fig. 1, all patterns a few 100 nm deep. The

Results and discussion

Both tri-layer nanoimprint lithography and direct embossing of plastic sheets result in well-defined nanostructure arrays. Fig. 2 displays two examples of the replicated nanopillar arrays, obtained by nanoimprint lithography (a) and plastic imprinting (b). They were taken from a large area of 320-nm period triangular array of 150-nm diameter pillars of 200-nm height, placed inside a 200-nm deep and 50-μm wide channel and extending over a total length of 7 mm. Excellent homogeneity over the

Conclusion

We have demonstrated the versatility of using either nanoimprint or nano-embossing in microfluidic device fabrication. High density nanostructures can be easily replicated and integrated into more complex microfluidic systems. The assembled structures, either a patterned silicon wafer covered by a PDMS plate or an all-plastic device, are reliable, low-cost, and should allow a number of bio-applications.

Acknowledgments

The authors would like to thank Y. Baba of the University of Tokushima, Japan, for useful discussions.

References (7)

  • Y. Chen et al.

    Eur. Phys. J. Appl. Phys.

    (2000)
  • M. Ueda et al.

    Electrophoresis

    (2000)
  • S.Y. Chou et al.

    Appl. Phys. Lett.

    (1995)
There are more references available in the full text version of this article.

Cited by (0)

View full text