Summary thesis Jeroen Haneveld

This thesis deals with the fabrication and characterization of nanochannels (channels with at least one dimension in the sub-100 nm range). These channels are important for various areas of research, including DNA analysis systems and chemical sensors. In addition, the behavior of liquids in nano-confinement is of interest for many of the applications. The technologies currently used to fabricate nanochannels are often expensive and/or time consuming or simply not accurate enough. This creates a need for controlled, yet simple fabrication technologies.

Two possible methods were developed to fabricate nanochannels with a height below 100 nm. The first one is based on wet anisotropic etching of silicon, giving nanochannels with rectangular cross-sections. When extremely low depth of the channels is necessary (down to 5 nm), the use of a patterned silicon dioxide layer is advisable. Both technologies show excellent surface roughness and good uniformity. Sealing of the channels was achieved by direct bonding of the wafers to silicon or Borofloat glass wafers. In addition, complete chips including nanochannels and integrated fluidic reservoirs, measurement rulers and power blasted access holes were designed and fabricated. The nanochannels on the chips have depths ranging from 5 to 150 nm. To study the behavior of fluids in nano-confinement, the chips were filled with various liquids (water, sodium chloride solution and cyclohexane). The filling speed was compared to a theoretical model, based on Washburn’s equation for capillary filling. Qualitative agreement with the model was confirmed, but quantitatively a reduced filling speed of the liquids was observed, the reduction becoming larger when the channel depth decreased.

In addition, two technologies were described to enable the fabrication of two-dimensional nanochannels, which have both a depth and a width in the nanometer range. Laser interference lithography, in combination with wet anisotropic etching of silicon is able to directly create such nanochannels. Local oxidation of silicon edges can provide very thin (sub-10 nm) nano-ridges, which can be transferred to a silicon wafer by imprint lithography.

In conclusion, the future fabrication of nanochannels can be greatly simplified when using the technology described in this thesis. Further research remains to be done on the characterization of fluid flow inside the channels.